Applied Computer Science: MAT 1102

INTRODUCTION TO
STATISTICS AND
PROBABILITY
Nafy Aidara
 Introduction to Statistics and Probability

Foreword
The African Virtual University (AVU) is proud to participate in increasing access to education in
African countries through the production of quality learning materials. We are also proud to
contribute to global knowledge as our Open Educational Resources are mostly accessed from
outside the African continent.

This module was developed as part of a diploma and degree program in Applied Computer
Science, in collaboration with 18 African partner institutions from 16 countries. A total of 156
modules were developed or translated to ensure availability in English, French and Portuguese.
These modules have also been made available as open education resources (OER) on oer.avu.
org.

On behalf of the African Virtual University and our patron, our partner institutions, the African
Development Bank, I invite you to use this module in your institution, for your own education,
to share it as widely as possible and to participate actively in the AVU communities of practice
of your interest. We are committed to be on the frontline of developing and sharing Open
Educational Resources.

The African Virtual University (AVU) is a Pan African Intergovernmental Organization established
by charter with the mandate of significantly increasing access to quality higher education and
training through the innovative use of information communication technologies. A Charter,
establishing the AVU as an Intergovernmental Organization, has been signed so far by
nineteen (19) African Governments - Kenya, Senegal, Mauritania, Mali, Cote d’Ivoire, Tanzania,
Mozambique, Democratic Republic of Congo, Benin, Ghana, Republic of Guinea, Burkina Faso,
Niger, South Sudan, Sudan, The Gambia, Guinea-Bissau, Ethiopia and Cape Verde.

The following institutions participated in the Applied Computer Science Program: (1) Université
d’Abomey Calavi in Benin; (2) Université de Ougagadougou in Burkina Faso; (3) Université
Lumière de Bujumbura in Burundi; (4) Université de Douala in Cameroon; (5) Université de
Nouakchott in Mauritania; (6) Université Gaston Berger in Senegal; (7) Université des Sciences,
des Techniques et Technologies de Bamako in Mali (8) Ghana Institute of Management and
Public Administration; (9) Kwame Nkrumah University of Science and Technology in Ghana; (10)
Kenyatta University in Kenya; (11) Egerton University in Kenya; (12) Addis Ababa University in
Ethiopia (13) University of Rwanda; (14) University of Dar es Salaam in Tanzania; (15) Universite
Abdou Moumouni de Niamey in Niger; (16) Université Cheikh Anta Diop in Senegal; (17)
Universidade Pedagógica in Mozambique; and (18) The University of the Gambia in The
Gambia.

Bakary Diallo

The Rector

African Virtual University

2
Production Credits
Author
Nafy Aidara

Peer Reviewer
Robert Oboko

AVU - Academic Coordination

Dr. Marilena Cabral

Overall Coordinator Applied Computer Science Program

Prof Tim Mwololo Waema

Module Coordinator
Florence Tushabe

Instructional Designers
Elizabeth Mbasu

Benta Ochola

Diana Tuel

Media Team
Sidney McGregor Michal Abigael Koyier

Barry Savala Mercy Tabi Ojwang

Edwin Kiprono Josiah Mutsogu

Kelvin Muriithi Kefa Murimi

Victor Oluoch Otieno Gerisson Mulongo

3
 Introduction to Statistics and Probability

Copyright Notice
This document is published under the conditions of the Creative Commons
http://en.wikipedia.org/wiki/Creative_Commons

Attribution http://creativecommons.org/licenses/by/2.5/

Module Template is copyright African Virtual University licensed under a Creative Commons
Attribution-ShareAlike 4.0 International License. CC-BY, SA

Supported By

AVU Multinational Project II funded by the African Development Bank.

4
Table of Contents
Foreword 2

Production Credits 3

Copyright Notice 4

Supported By 4

Course Overview 8

Prerequisites. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8

Materials. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8

Course Goals. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8

Units. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9

Assessment . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9

Schedule. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10

Readings and Other Resources . . . . . . . . . . . . . . . . . . . . . . . . . . . 10

Unit 0: Pre-Assessment 12

Unit Introduction. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12

Unit Objectives . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12

Key Terms. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13

Unit Assessment. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 15

Grading Scheme. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 22

Unit Readings and Other Resources. . . . . . . . . . . . . . . . . . . . . . . . 22

Unit 1: Basic Statistics and its Application in ACS 23

Unit Introduction. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 23

Unit Objectives . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 23

Key Terms . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 23

Learning Activities. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 25

Activity 1 statistics 25

Activity Details 25

Activity 2 Types of statistical data 31

Activity Details 31

5
 Introduction to Statistics and Probability

Activity 3 Tabular and graphic representation 35

Activity Details 35

Unit Summary. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 59

Unit Assessment. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 59

Unit 3: Linear Regression 64

Introduction to Unit . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 64

Unit Goals . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 64

Key Terms . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 64

Learning activities . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 65

Activity 1 The simple linear regression model 65

Introduction 65

Details of the activity 65

Activity 2 Least squares 77

Unit Summary. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 85

Unit Assessment. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 85

Unit 4: Applications of Probability and Statistics in ACS 88

Unit Introduction. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 88

Unit Objectives . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 88

Key Terms . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 88

Learning Activities. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 89

Activity 1 Naive Bayesian and k-means neighbor 89

Activity Details 90

Activity 2 Decision Tree for classification 97

Activity Details 98

Activity 3 Clustering 101

Activity Details 101

Unit Summary. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 107

Unit Assessment. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 108

Grading Scheme. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 112

Answer. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 112

6
Unit Readings and Other Resources. . . . . . . . . . . . . . . . . . . . . . . . 113

Module Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 113

Module Course Assessment . . . . . . . . . . . . . . . . . . . . . . . . . . . . 114

Course References. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 117

7
 Introduction to Statistics and Probability

Course Overview
Welcome to Introduction to Probability and Statistics
The aim of the course is to equip students with basic knowledge in probability and statistics
needed for their studies in ACS. In modern computer science, software engineering, and
other fields, the need arises to make decisions under uncertainty. Probability and Statistics
helps computer science students solve problems and make decisions in uncertain conditions,
compute probabilities and forecasts, and evaluate performance of computer systems and
networks. At the end of the course, students should be able to apply Probability & Statistics
in the context of ACS. Indeed, they will be able to use statistical concepts, probabilistic
calculations, methods of observations, sampling techniques, analysis and classification of
variables in interpreting data, and inferring design variables results.

Prerequisites
Basic Mathematics and Calculus, Basic IT skills (spreadsheet)

Materials
The materials required to complete this course are:

• DeGroot, Morris H., and Mark J. Schervish. Probability and Statistics. 3rd ed.
Boston, MA: Addison-Wesley, 2002. ISBN: 0201524880.
• Lecture notes, calculators, computers and Internet connectivity

Course Goals
Upon completion of this course the learner should be able to:

• Analyze counting problems in computer science

• Apply basic concepts of probability;
• Apply random situations involving the concept of chance;
• Design studies using descriptive statistical methods to interpret and analyze the
results from these formulating conjectures;
• Select appropriate statistical data analysis methods;
• Interpret information of a statistical nature;

8
 Course Overview

Units
Unit 0: Pre-Assessment

This unit will help students assess their level of knowledge in probability and statistics and also
evaluate their competencies in basic mathematics. It is not compulsory but could serve as a
guide for both the teachers and the students to identify the knowledge gaps.

Unit 1: Basic Statistics and introduction to SPSS

This unit will introduce the student to the basic concepts of statistics and the use of statistical
software in computer science problems solving.

Unit 2: Basic Probability and its applications in ACS

This unit will introduce students to the basic concepts in probability.

Unit 3: Linear Regression and Its Applications in ACS

Unit 4: Practical Applications of probability and statistics in ACS

Specific examples in applied computer science will be provided in this unit.

Assessment
Formative assessments, used to check learner progress, are included in each unit.

Summative assessments, such as final tests and assignments, are provided at the end of each
module and cover knowledge and skills from the entire module.

Summative assessments are administered at the discretion of the institution offering the course.
The suggested assessment plan is as follows:

1 Assignments 10%

2 Quizzes 10%

3 Tests 30%

4 Final Examination 50%

9
 Introduction to Statistics and Probability

Schedule

Unit Activities Estimated time

Unit 1 1.     Learning Activity 1 10 hours

2.     Learning Activity 2 10 hours

3.     Learning Activity 3 10 Hours

Unit 2 4.     Learning Activity 1 10 hours

5.     Learning Activity 2 10 hours

6.     Learning Activity 3 10 Hours

Unit 3 7.     Learning Activity 1 10 hours

8.     Learning Activity 2 10 hours

9.     Learning Activity 3 10 Hours

Unit 4 10. Learning Activity 1 10 hours

11. Learning Activity 2 10 hours

12. Learning Activity 3 10 Hours

Readings and Other Resources

The readings and other resources in this course are:

Unit 0

Required readings and other resources:

• DeGroot, Morris H., and Mark J. Schervish. Probability and Statistics. 3rd ed.
Boston, MA: Addison-Wesley, 2002. ISBN: 0201524880.
• John A. Rice, Mathematical Statistics and Data Analysis (with CD Data Sets)
(Duxbury Advanced). 3rd Edition, Cengage Learning, 2006, ISBN-13 978-
0534399429 for reading.
• http://ocw.mit.edu/courses/mathematics/18-05-introduction-to-probability-and-
statistics-spring-2005/index.htm for self practicing. Lots of interesting exercises.
• http://books.google.com/books/about/
Probability_and_Random_Processes_With_Ap.htm

10
 Course Overview

Optional readings and other resources:

• http://mathworld.wolfram.com/Probability: Wolfram is a useful site that provides

insights in number theory while providing new challenges and methodology in
number theory.
• http://en.wikipedia.org/wiki/Probability: Mathsguru is a website that helps
learners to understand various branches of number theory module. It is easy to
access through Google search and provides very detailed information on various
probability questions.

Unit 1

Required readings and other resources:

• DeGroot, Morris H., and Mark J. Schervish. Probability and Statistics. 3rd ed.
Boston, MA: Addison-Wesley, 2002. ISBN: 0201524880.
• John A. Rice, Mathematical Statistics and Data Analysis (with CD Data Sets)
(Duxbury Advanced). 3rd Edition, Cengage Learning, 2006, ISBN-13 978-
0534399429 for reading.
• http://ocw.mit.edu/courses/mathematics/18-05-introduction-to-probability-and-
statistics-spring-2005/index.htm for self practicing. Lots of interesting exercises.
• http://books.google.com/books/about/
Probability_and_Random_Processes_With_Ap.htm

Optional readings and other resources:

• http://mathworld.wolfram.com/Probability: Wolfram is a useful site that provides

insights in number theory while providing new challenges and methodology in
number theory.
• http://en.wikipedia.org/wiki/Probability: Mathsguru is a website that helps
learners to understand various branches of number theory module. It is easy to
access through Google search and provides very detailed information on various
probability questions.

Unit 2

Optional readings and other resources:

• http://mathworld.wolfram.com/Probability: Wolfram is a useful site that provides

insights in number theory while providing new challenges and methodology in
number theory.
• http://en.wikipedia.org/wiki/Probability: Mathsguru is a website that helps
learners to understand various branches of number theory module. It is easy to
access through Google search and provides very detailed information on various
probability questions.

11
 Introduction to Statistics and Probability

Unit 0: Pre-Assessment
Unit Introduction
The purpose of this unit is to determine your grasp of knowledge related to this course.
Students will be able to assess their understanding in basic algebra and arithmetic prior to the
beginning of the course. Students evaluate their basic skills in calculus and IT.

Unit Objectives
Upon completion of this unit you should be able to:

• Define the basic concepts in arithmetic

• Draw graph from a specific data set
• Analyse and Interpret data.
• Draw conclusions from the interpretations.

12
 Unit 0: Pre-Assessment

Key Terms
Probability: provides mathematical models for random
phenomena and experiments, such as: gambling, stock
market, packet transmission in networks, electron
emission, noise, statistical mechanics, etc.

Statistics: Statistics is a very broad subject, with

applications in a vast number of different fields. In
generally one can say that statistics is the methodology
for collecting, analyzing, interpreting and drawing
conclusions from information.

Population: A (statistical) population is the set of

measurements (or record of some qualitative trait)
corresponding to the entire collection of units for which
inferences are to be made.

Sample: A sample from statistical population is the set

of measurements that are actually collected in the course
of an investigation.

Parameter: A parameter is an unknown numerical

summary of the population. A statistic is a known
numerical summary of the sample which can be used to
make inference about parameters.

Probability: provides mathematical models for random

phenomena and experiments, such as: gambling, stock
market, packet transmission in networks, electron
emission, noise, statistical mechanics, etc.

Descriptive Statistics: deals with procedures used

to summarize the information contained in a set of
measurements.

Measures of central tendency: There are different ways to measure the central tendency of a
data set: Mean, Mode and Median.

Mean: It represents the average measure of the data set.

Median: It represents the middle value of the data set

Mode: This is the value that occurs the most frequently.

Variance: It is the nonnegative integer that measure the amount of variation within the data
set.

13
 Introduction to Statistics and Probability

Standard Deviation: The square root of the variance.

Percentiles: dividing the ordered data set by the use of ordinates so that the amount of data
points less than the ordinate is some percentage of the total amount of observations. The
values corresponding to such areas are called the percentiles.

Interquartile ranges: This is the difference between the first quartile and the third quartile.

Skewness: If the distribution of the data set ical about any value, it is said to be skewed. If
the data set has a few more lower values, it is said to be skewed to the left. If it ha more higher
values, it is said to be skewed to the right

Random variable: Suppose that to each sample point we assign a number. We then have
a function defined on the sample space. This function is called a random variable. It is usually
denoted by X or Y.

Discrete Random Variable: A random variable that takes on a finite or countably infinite
number of values is called a discrete random variable.

Continuous Random Variable: A random variable that takes on an uncountably infinite

number of values is called a non discrete/continuous random variable.

14
 Unit 0: Pre-Assessment

Unit Assessment
Check your understanding!

Diagnostic Test

Instructions

The following tests are intended to diagnose weaknesses that you might have in the areas
under the four units. This is a self-administered test. As you begin your studies on statistics
and probability, take this course pre-assessment to help you understand better what you
may already know about the subject and how best to study the course material. If you find
you do well on parts of this self-test, you can move more quickly over that subject in the
study guide and leave yourself more time for other sections. This approach will help you
determine which areas to review in order to be successful in this course. If you don’t do well
at first, do not worry. After taking the test you can check your answers against the given
answers and, if necessary, refresh your skills. You are not required to hand in your work.

For multiple choice and true/false questions, circle the best answer.

1. __________ is the science that involves collection, analysis, and interpretation of data in
order to make inferences about populations.

A. Probability

B. Statistics

C. Management

D. None of the above

2. Out of 100 numbers, 20 were 4s, 40 were 5s, 30 were 6s and the remainder were 7s. Find
the arithmetic mean of the numbers.

A. 0.22

B. 0.53

C. 2.20

D. 5.30

3. The following set of raw data shows the length, in millimeters, measured to the nearest
mm, of each of 40 leaves taken from plants of a certain species. This is the table of
frequency distribution.

15
 Introduction to Statistics and Probability

Length (mm) Frequency ( )

25 – 29 2

30 – 34 4

35 – 39 7

40 – 44 10

45 – 49 8

50 – 54 6

55 – 59 3

4. Find the mean of the distribution.

A. 2

B. 3

C. 4

D. 5

5. Find the mode of the following data: 5, 3, 6, 5, 4, 5, 2, 8, 6, 5, 4, 8, 3, 4, 5, 4, 8, 2, 5, and 4.

A. 4

B. 5

C. 6

D. 8

6. The range of the values a probability can assume is

A. From 0 to 1

B. From -1 to +1

C. From 1 to 100

D. From 0 to 5

16
 Unit 0: Pre-Assessment

7. The grades (on a scale of 100) of 69 students are depicted on a stem and leaf diagram.

Determine the median.

5 | 778

6 | 1122334

6 | 555666777799999

7 | 001123333344

7 | 5555666777779

8 | 00222244

8 | 5567788999

9|1

A. 5

B. 7

C. 9

D. 2

8. If two children are picked from a group of ten. Determine the probability of picking two
children

A. 0.2

B. 0.6

C. 0.4

D. 0.3

9. A box contains three coins with a head on both sides, four coins with a tail on both sides,
and two fair coins. If one of these nine coins is selected at random and tossed once, what is
the probability that a head will be obtained?

A. 3/9

B. 4/9

C. ½

D. 2/9

17
 Introduction to Statistics and Probability

10. A nation-wide professional qualifying exam has a mean score of 600 with a standard
deviation of 50. A random sample of 100 examinees was selected. The sample mean was
630. Determine the standardized value or z-score of the sample mean in this instance.

A. 0.6

B. 0.2

C. 0.45

D. 0.5

11. What type of graphical representation would be best to display the following data?
Leslie sold 47 hot dogs. Kelly sold 32 hot dogs. Jessie sold 30 hot dogs. Carlos sold 4 hot
dogs. You want to show a comparison among these people to determine who sold the
most hot dogs and who sold the least hot dogs.

A. Bar graph

B. Line graph

C. Stem and leaf plot

D. Histogram

12. Using the data below, approximately what percentage of students buys either hot dogs
or hamburgers from the cafeteria at lunch?

Food Bought Number of

Students

Hamburgers 241

Hotdogs 361

Pizzas 129

Salad 45

Sandwich 63

Nothing 84

Total 923

18
 Unit 0: Pre-Assessment

A. 65%

B. 26%

C. 39%

D. 75%

13. What type of graphical representation would be best to display the following data?

{ 3,5,8,5,2,34,8,9,16,21} You want to show which numbers in the following set are outliers
and you want to show the mean of the following set.

A. Line graph

B. Bar Graph

C. Stem and leaf plot

D. Line plot

14. If P(A)=½ and P(B)=⅓. Find P(AuB).

A. 1/6

B. 2/3

C. 5/6

D. 1/3

15. The mean age of 5 persons in a room is 30 years. A 36-year-old person walks in. What is
the mean age of the persons in the room now?

A. 35

B. 34

C. 31

D. 30

19
 Introduction to Statistics and Probability

16. 99.7% of the data falls within __________ standard deviations of the mean on the normal
distribution.

A. 1

B. 2

C. 3

D. 4

17. A z-value of __________ is used for a 90% confidence interval.

A. 1

B. 2

C. 1.5

D. 3

18. The normal distribution is a __________-shaped distribution.

A. Oval

B. Bell

C. Circ

D. None of the above

19. The table below gives the probability density function of a random variable X. Find the
expected value of X.

X Probability

-2 0.1

0.2

0 0.3

1 0.1

5 0.3

20
 Unit 0: Pre-Assessment

A. 0.3

B. 1.2

C. 2.5

D. None of the above

20. The loss due to a fire in a commercial building is modeled by a random variable X with
density function if and 0 elsewhere. Given that a fire loss exceeds 8, what is the probability
that it exceeds 16?

A. 1/25

B. ⅛

C. 1/9

D. ⅓

Answers:

1. B

2. C

3. A

4. B

5. D

6. A

7. B

8. B

9. C

10. B

11. A

12. D

13. D

14. C

15. A

21
 Introduction to Statistics and Probability

16. A

17. B

18. A

19. C

20. C

Grading Scheme
Each correct answer will carry one (1) mark.

Unit Readings and Other Resources

The readings in this unit are to be found at the course-level section “Readings and Other
Resources”.

22
 Unit 1: Basic Statistics and its Application in ACS

Unit 1: Basic Statistics and its

Application in ACS
Unit Introduction
This unit provides a set of concepts and methods that are designed to enable students to be
able, through correct application, to interpret and analyze sample data, apply the appropriate
techniques for data analysis and critically interpret the results.

Unit Objectives
Upon completion of this unit you should be able to:

1. The concept of Statistics and its divisions;

2. Identify the concept of population, sample and random experiment;

3. Describe the organization and type of statistical data

4. Identify the different types of sampling

5. Compute the mean, mode and the measures of variation

Key Terms
Statistics: The branch of mathematics that investigates
the processes of obtaining, organizing and analyzing
data on a population or on a collection of all beings, and
the methods to draw conclusions and make inferences or
predictions based on these data.

Descriptive Statistics: Responsible for the organization

and description of the information

Inductive and Inferential Statistics: Understands

the generalization process, from the analysis and
interpretation of sample data.

Population or Universe: fundamental set of all elements

with at least one common feature.

Sample or Event: Subset of a universe.

23
 Introduction to Statistics and Probability

Experiment Random or not deterministic: Statistical

Observation of a any phenomenon such as the toss of a
coin (not addicted) to observe the end result, heads or
tails.

Variables: It is a symbol representing certain characteristic

of a population or sample. In other words, a variable is
a characteristic of the population can be measured
according to some scale.

Quantitative Variables: are the characteristics that can be

measured in a quantitative scale that is, having numerical
valuesthat make sense. It can be continuous or discrete.

Discrete Variables: measurable characteristics that can

assume only a finite or infinite number of values countable
and thus make sense only integer values. They are usually
the result from counting. Examples: number of children,
number of bacteria per liter of milk, number of cigarettes
smoked per day.

Continuous variables, measurable characteristics that take

values on a continuous scale (on the real line), for which
fractional values make sense. Usually should be measured
using an instrument. Examples: weight (balance), height
(ruler), time (clock ), blood pressure , age .

Qualitative Variables (or categorical): features do

not have quantitative values , but rather , are defined
by several categories , i.e. , represent a classification of
individuals. It can be nominal or ordinal.

Nominal variables: there is no ordering among the

categories. Examples: sex, eye color, smoker / non-smoker,
patient / healthy.

Ordinal variables: there is an ordering between the

categories. Examples: education ( 1st, 2nd , 3rd degree )
, disease stage ( initial, intermediate , terminal) , month of
observation (January, February , ..., December.

Sampling: set of procedures by which you select a sample

of a population.

24
 Unit 1: Basic Statistics and its Application in ACS

Sampling probabilistic - Procedure in which all elements

of the population have a known and a probability of zero
to integrate sample.

Not probabilistic sampling Intentional Sampling -

Sampling Probabilistic not subject to specific objectives
of the investigator.

Sampling unintentional - Sampling Probabilistic not

governed by criteria of convenience and / or availability
of respondents.

Raw data: set of data that have not been organized

numerically obtained after the critical values.

Role: It is an arrangement of raw data in ascending order.

Total Width (AT): the difference between the highest and

the lowest value observed.

Learning Activities

Activity 1 statistics

Introduction

Every day we are exposed to a large amount of numerical information. Depending on the
situation, why are consumers of numerical information, now we need to produce them. Thus,
we need knowledge and training to understand the construct them. Procedures, techniques
and statistical methods are fundamental to aid the implementation of these tasks. In summary,
the statistic is SCIENCE The data, a science for the producer and the consumer of numerical
information. It involves collection, classification, summarization, organization, analysis and
interpretation of data.

Activity Details

Statistics is a part of applied mathematics that provides methods for collection , organization,
presentation , analysis and interpretation of data.

Statistics is divided into two areas:

• Descriptive Statistics is part of the statistic that takes care of the collection ,
organization and description of the observed data;
• Inferential Statistics (Inductive Statistics) is the part of the statistic that tries to
generalize findings to a population from the analysis and interpretation of data
from a sample.

25
 Introduction to Statistics and Probability

The statistic is present in many activities that directly affect our lives, for example:

• The analysis of traffic problems;

• The study of the effects of various drugs;
• Quality control of products;
• The evaluation of teaching techniques ;
• The behavior;
• The study

The statistical scope has expanded considerably. One reason is the increasing use of statistics
in various sectors such as agriculture, education, politics, ENGINEERING, psychology,
economics and administration. Another reason why the development of statistics in recent
years is the technological developments that facilitated our ability to handle information

The analysis of a statistical problem is done over several phases:

26
 Unit 1: Basic Statistics and its Application in ACS

i. Problem Definition: Know exactly what you want to search; the purpose of establishing the
analysis and definition of the population.

ii. Sampling and Data Collection: Operational phase. The process of selection and systematic
recording of data, with a particular purpose. Data can be primary (published by the person
or organization) or secondary (when they are published by another organization).

iii. Treatment and Presentation of Data: Summary of data through its counting and grouping. It
is the data classification, using tables or graphics.

iv. Analysis and Data Interpretation: The last phase of statistical work is the most important
and delicate. Is essentially linked to the calculation of measures and coefficients, whose
main purpose is to describe the phenomenon of behavior under study (descriptive
statistics). In inductive statistical data interpretation are based on probability theory.

Conclusion

This is an introductory activity based on the basic concepts of statistics, for the same
TRANSMISSION trainees must carry out various tasks such as:

1st Activity: Reading the references mandatory;

2nd Activity: Investigation of different concepts on the Internet ;

3rd Activity: Group work - Prepare a summary of 50 words ,

explaining the circumstances in which sampling is preferable to a
census (all members of the public ) ?. Exemplifies.

Activity 4: application exercises Resolution.

Practical Exercise

1. Choose the correct alternative:

a. Population or universe is:

I. Set people.

ii. Individuals presenting a special feature.

iii. Set all individuals with a common characteristic object of study.

b. The variable is discrete when:

i . Given two real values , we can find at least one value among them .

ii. Given two real values, we cannot find values between them.

iii. Given two real values, the difference between them is zero.

27
 Introduction to Statistics and Probability

c. The main stages of the statistical method are:

I . Data collection, sampling, tabular presentation and layout and definition of

problems.

ii. Sampling, tabular presentation, verification of data, interpretation of data

and planning.

iii. Problem definition, planning, data collection, calculation, data presentation,

analysis and interpretation of data.

d. Part of the population removed to analyze it is called:

I. Universe;

ii. Party;

iii. Piece;

iv. Raw Data;

v. Sample.

2. The intention was to make a study of the number of siblings of students in the 10th grade of
a secondary school.

For this, a survey was carried out to which 60 students answered. Indicate:

a) The study population

b) The chosen sample;

c) The study variable and rate it

3.In a survey about the time (in hours) that Guineans are connected to the internet, were
interviewed 2500 people.

Assuming, in Bissau city there are about 400 000 people, identifying the population and the
sample in this situation.

4. The director of a college, in which they are enrolled 280 boys and 320 girls, wanting to know
the conditions of extra- school life of their students and not having time to interview all families,
decided to do a survey by sampling 10% of these clients. Get, this director, the elements of the
sample.

5. A city X has the following table of his faculties:

28
 Unit 1: Basic Statistics and its Application in ACS

SCHOOL NUMBER OF STUDENTS

MALE FEMALE

A 80 95

B 102 120

C 110 92

D 134 228

E 150 130

F 300 290

Total 876 955

Get a stratified proportional sample of 120 students.

6. A population is divided into three levels, with sizes , respectively , n1 = 40 , n2 = n3 = 100

and 60. Knowing that, when held a proportionate stratified sampling, nine elements of the
sample were taken from the 3rd layer, determine the total number of sample elements.

7. Show how could a sample of 32 elements of a population consisting of 2,432 ordered

elements.

In the general ranking, which of the following elements would be chosen to belong to the
sample, given that the element of order 1420 belongs to it ?

1.648º, 290 º, 725º, 2.025º, 1.120º .

8. Identify which of these sample types are used : random, systematic, convenience, stratified
or cluster

a.News on TV - A news reporter of the Globe network analyzes the

reaction to an impressive history interviewing people passing in
front of his studio.

b. Telephone surveys - In a survey on the operation of the 1059 MTN

people, the subjects of the interview were selected using a computer
to randomly generate telephone numbers, which were then dialed.

c. Car ownership - a researcher at General Motors divided all cars

registered in categories of subcompact, compact , medium ,
intermediate and large . He’s searching 200 car owners in each
category.

d. Drink between Students - Motivated by the fact that a student has

died from excessive drinking, college did a study of student drinking
habit, randomly selecting 10 different classes and interviewing all
the students in each class.

29
 Introduction to Statistics and Probability

e. Of Sobriety Checkpoint - The author was an observer from one

point police sobriety checks, in which every fifth driver was stopped
and interviewed. (He witnessed the arrest of a former student).

f. Urn of Boca Research - A news network is planning a survey in which

100 polling stations will be selected at random and all voters will be
interviewed on leaving the site.

Anthropometry

g. A student obtains statistical data on height / weight interviewing

family members.

h. Medical Research - A researcher at ENA examines all heart patients

from each of the 30 hospitals randomly selected.

9. In a binder with 500 numbered chips and ordered from 1 to 500, select 10 records for a
search.

Answers

1a) iii b) ii c) iii d) v

Sex Number of Rate of Number of

students sampling selected
fraction students

Male 876 0.07 57.41

Female 955 0.07 62.59

Total 120

6 – 30

7 - 1.648º

8. Note that k = 500/10 = 50. A record draws between 1 and 50, for example, the plug 17
will be selected the next number 17 + 50 = 67, and so forth, and therefore, the sample is
composed of records:

17, 67, 117, 167, 217, 267, 317, 367, 417, 467

30
 Unit 1: Basic Statistics and its Application in ACS

Activity 2 Types of statistical data

Introduction

Statistical data are the submission of observations of a particular variable, either qualitative
or quantitative nature, in order to describe the entire set of units observed in summary
or summary form. Statistical data form the basis of all study and statistical analysis of the
discussed context, give that the statistical data are considered as the primary ingredient to
any investigation. Statistical data can be collected from existing sources or collected through
survey and experimental studies; they can be of different types and therefore need to be
treated with different statistical methods.

Activity Details

The description and interpretation of data is an essential part of statistics. The quality of
the solution of the statistical problem is directly related to the quality of the data obtained.
Therefore, in terms of utilization, appropriate methods for data collection depend to the
problem to be studied.

Figure representing the types of statistical data.

Conclusion

Exercises: Application and concentration

1. In a study in a school, data was collected for the following variables:

(a) age (E) time spent daily in study

(b) grade (F) distance from home to school

(c) sex (G) study site

31
 Introduction to Statistics and Probability

(d) note in the discipline of Mathematics (H) number of siblings

a)The indicated variables, which are quantitative and which are qualitative?

b) Of quantitative variables, which are continuous says.

2. Rate the variable as qualitative (nominal or ordinal) or quantitative (discrete or continuous):

a. Population: members of a club.

b. Variable: height (in meters).

c. Population: cars produced by an automaker.

d. Variable: number of ports.

e. Population: animals in a zoo.

f. Variable: predominant color.

g. Population: candidates for a job opening.

h. Variable: Education.

i. Population: soccer players in a club.

j. Variable: position in which they operate.

k. Population: recipes in a cookbook.

l. Variable: preparation time.

3. a network of shops are doing a survey on customer satisfaction. One of the questions that
the client must answer is: “Are you satisfied with our service.” The categories of responses
were.

32
 Unit 1: Basic Statistics and its Application in ACS

A sample of 50 people who answered the questionnaire, provided the following answers to
specific question (to help the procedures of the results via computer a numerical scale was
used where: 2 = satisfied, 1 = dissatisfied, 0 = undecided)

These data are quantitative or qualitative. Review your answer.

4 . A survey of a newspaper surveyed 2500 adults, “Are you satisfied with the economic
situation in the country today.” The response categories were dissatisfied, satisfied and
indecisive.

a. What is the size of the survey sample?

b. The data collected were qualitative or quantitative?

c. Of those who responded, 28% said they were dissatisfied with

the economic situation. How many individuals have provided this
answer?

5 .Classify the following variables as qualitative and quantitative and the latter is discrete or
continuous:

a) No. passengers on the bus from Bafatá line;

b) Education of a group of people;

c) The average weight of newborns from maternity;

d) Altitude above sea level;

e) A survey conducted with 1,015 people indicates that 40 of them are

subscribers to a broadband Internet service;

f) The electronic indicates that the player radar last snapped ball
82,3mi / h;

g) The time spent for a person to make a Bissau drive to Gabu is

approximately 2: 40h at an average speed of 100km / h.

h) The students eye color;

i) Production of cashew nuts in Guinea-Bissau;

j) Number of defects on TV equipment;

k) The point obtained in each play of a given

33
 Introduction to Statistics and Probability

Answers

1.a. Quantitative (A), (D), (E), (F), (H)

b. Qualitative: (B) (C) (G)

c. Are continuous quantitative variables (E), (F) and optionally (A); the variable Age is also
continuous, it can take any value in a range, although it is usually treated as discrete)

2. a) continuous Quantitative;

b) Quantitative discreet;

c) Qualitative ordinal;

d) Qualitative

ordinal;

e) nominal Quantitative;

f) Continuous Quantitative

3. Are quantitative were converted into numerical scale according to a number

4.a. 2 500 adults

b. qualitative

c. 700 adults

5. a. Quantitative discrete

b. qualitative

c. quantitative continuous

d. quantitative continuous

e. quantitative discrete

f. qualitative

g. quantitative continuous

h. qualitative

i. quantitative continuous

j. quantitative continuous

k. quantitative discrete

34
 Unit 1: Basic Statistics and its Application in ACS

Activity 3 Tabular and graphic representation

Introduction

Descriptive statistics, whose basic objective is to synthesize a series of related values, thereby
enabling to have a global view of the variation of these values, organizes and describes the
data in three ways: through tables, graphs and descriptive measures.

The table is a representation summarizing a set of observations, while the graphics are
presentations of data, whose goal is to produce a faster and more vivid impression of the
phenomenon under study.

Activity Details

Table

It is very common these days due to the use of computers, conduct research where data
collection results in large collections (quantity) of data for analysis and it becomes almost
impossible to understand them, as to (s) particular (s) objective (s) of study if these data are not
summarized. In other words, the manner in which data were collected do not allow an easy and
quick way, which extract information.

Table of Elements

Every table should be simple, clear, objective and self-explanatory

The provision of a table can be generalized as shown in the Figure below.

Table X Title answering the questions: what, where and when?

indicator column Header

Online content Cell

Source: Data source

Note: Insightful information table Body

It is noteworthy that the tables must be numbered in ascending order or in which they appear
in the text, as in the case of scientific papers; the top and bottom edges must be closed with
horizontal lines while the left and right not, or cannot be closed by vertical lines separating the
columns in the body of the table. It is also necessary that the number of decimal places to be
standardized.

35
 Introduction to Statistics and Probability

Frequency Distribution

When studying a mass of data is frequent interest summarize the variable information.

Steps for building a frequency distribution:

1. Find the values that can be assumed by the variable-;

2. Arrange the values in ascending order, in the left column of your table;

3. Make a number of times each value appears consolidation;

4. Enter the numbers found in step 3 in the column next to “values” in the column named
“Frequency”;

Relative frequency or percentage

36
 Unit 1: Basic Statistics and its Application in ACS

Cumulative absolute frequency, denoted by Fai. These frequencies are obtained by adding
the absolute frequency of the value considered, the previous absolute frequencies to the same
value.

Cumulative relative frequency

Distribution in classes or interval

“The distribution of frequencies of classes is suitable for quantitative provide continuous or

discrete data with a large number of possible values”

It is necessary to divide the data into intervals or ranges of values that are

called classes. A class is a line of frequency distribution. The lowest class is called lower
bound (II) and the highest value of the class is called upper limit (Li). The range or class can be
represented the following ways:

a. li | ____ Li,, where the lower limit of the class is included in the absolute frequency count but
not higher;

b. read ____ | Li, where the upper limit of the class is included in the count but not less;

c. li | ____ | Li, where both the lower limit and the upper are included in the count;

d. ____ Li li, where the limits are not part of the count.

Methods for determining the number of classes

Rule Sturge

Square Rule or square root

where:

K= number of classes

N= total number of observations

Example

Age of students attending the course Statistical Inference, a course in Statistics at the
University of Amilcar Cabral Bissau, 03.21.2014.

37
 Introduction to Statistics and Probability

Range=37-18=19 years

AGE Fi

18 2

19 1

20 6

21 2

22 1

23 1

24 1

25 3

26 1

29 1

30 1

35 1

37 1

Total 22

Range=37-18=19 years

Age Xi fi fi % Fan fi%

18 |---22 20 11 0,5 11 0,5

22 |---26 24 6 0,2727 17 0,7727

26 |---30 28 2 0,0909 19 0,8636

30 |---34 32 1 0,0455 20 0,9091

34 |---38 36 2 0,0909 22 1

Total 22 1

38
 Unit 1: Basic Statistics and its Application in ACS

Types of frequencies curves

1.4 - Statistical Measures

Another way to summarize the data in a quantitative variable, plus tables and graphs, is
present them in the form of numeric values, called descriptive measures. These measures are
calculated from population data, are called parameters and calculated from sample data are
called estimators or statistics.

The descriptive measures are: measures position (measure of central tendency), measures of
dispersion, skewness and kurtosis measures.

1. Arithmetic mean

• Mostly used measure of central tendency;

• Is defined as the sum of the expected values of all observations (observation is an
element of a sample) divided by the number of observations;
• The symbol μ (mi) will be used to denote mean of a population;
• The symbol is used to denote the mean of a sample;

39
 Introduction to Statistics and Probability

Example: A sample of 3 newborns weight: 2,75kg, 3,25kg and 3.80 kg.

Here, n, the size of the sample is equal to 3. x1, first observation is 2,75kg; x2, second
observation is 3,25kg; x3 is 3,80kg.

= (2.75 + 3.25 + 3.80) = 9.80 / 3 = 3.27, that is, average weight is 3.27 kg.

Weighted arithmetic mean

40
 Unit 1: Basic Statistics and its Application in ACS

2. Median Arithmetic. Grouped Data

Calculate the average height of infants according to the table below.

Where:

Pm: Midpoint of classes

i: Simple Absolute frequency

Example: Calculate the average height of infants according to the table below.

Median (md)

Other measure used to indicate the center of a distribution;

Ordered elements in the sample, the median is the value (or not the sample belongs) which
divides in half, i.e. 50% of the sample elements are less than or equal to the median and 50%
are greater than or equal to the median

Example: Given the variable x = {1, 3, 0, 2.4}, the mean is 2.

To calculate the average median of a data set is due to:

41
 Introduction to Statistics and Probability

1) order the set; in the above example: x = {0,1,2,3,4}

2) Verify that there is an odd or even number of values in the set; in the above example: 5

Observations - odd

3) it is odd the median is the value that occupies the central position and is even will be the
average of the two central positions.

Li class = lower limit

N = total number of elements

Fant = cumulative frequency

= frequency of previous class

= amplitude of the class

Example:

• Create class
• Frequency Table (accumulated / Relative)
• Media, Mode, median

42
 Unit 1: Basic Statistics and its Application in ACS

Frequency table

Median: to get the result the average is necessary to add part-mean (x), and then divided by
the sum frequency (f)

M = 3016 = 149.6

Median: to find the result you must use the formula of the precise calculation of the median.

Mode

Mode is the value that appears most frequently in a distribution.

Examples:

Let x = {0 1 0 2 3 4 4 0 3 2 5 6}, Mode is 0.

Let x = {3 1 2 3 3 4 5 1.5 2 1.5 0 4 1.5 1.5 6} mode is 1.5. The series is unimodal.

Let X {2, 3, 4, 4, 4, 5, 6, 7, 7, 7, 8, 9} has two modes: 4 and 7. The series is bimodal.

Mode for grouped data in class

In order to calculate the mode of grouped data, you need to:

• Find the modal class. The modal class is the class interval that has the largest
frequency.
• Find the lower class boundary of the modal class ( )
• Find the difference of frequency between the modal class to its upper class (
).

43
 Introduction to Statistics and Probability

• Find the difference of frequency between the modal class to its lower class ( ).

Add the to products by , then add it to .

Formula Method

Mode=

• Lb= lower limit of the modal class

• b= frequency of the modal class - frequency of previous class to the modal class
• a = frequency of the modal class - frequency of posterior class to the modal class
• C = amplitude of the modal class

Separatrices

The series are equally separated. These measures are - the quartiles, deciles and percentiles.

Quartiles

We call quartiles the values of a series that fall into four (4) equal parts.

Three quartiles are therefore identified (Q1, Q2 and Q3) to divide the series into four equal
parts.

Note: The quartile 2 (Q2) will always be equal to the median of the series.

Example: Calculate the quartiles of the series: {5, 2, 6, 9, 10, 13, 15}

1. The first step to be taken is the sort (ascending or descending) of the values:

2. {2, 5, 6, 9, 10, 13, 15}

3. The value that divides the above series into two equal parts is greater than
9, then the Median = Q2 = 9 that will be.

44
 Unit 1: Basic Statistics and its Application in ACS

4. The steps for determining the Q1 of a set are as follows:

• Determine on the (adding column fi);

• Calculate the value of (n / 4) (whether n is even or odd!);
• To build the college column;
• To compare the value of (n / 4) with the college’s values, starting from the first
college class (the top!) And asking the question: “This college is greater than or
equal to (n / 4)? “ If the answer is NO, the college spent the next class. When the
answer is YES, we will stop and try the corresponding class! This will be our Class
of First Quartile.
• Finally, we will apply the formula for Q1, extracting the data Q1 of this class, we
just found! Again the formula:

Example: For the set below, determine the value of the third quartile!

0 - 10 2

20-Oct 5

20 - 30 8

30 - 40 6

40 - 50 3

n = 24

Step 1) We will find and calculate n (3n / 4):

Hence, we find that n = 24 and therefore (3n / 4) = 18

Step 2) builds the college:

45
 Introduction to Statistics and Probability

0 – 10 2 2

10 – 20 5 7

20 – 30 8 15

30 – 40 6 21

40 – 50 3 24

n = 24

Step 3) we compared the college values with the value of (3n / 4), asking the question of
practice, adapted to the third quartile:

0 – 10 2 2 → 2 is greater than or equal

to 18? DO NOT!

10 – 20 5 7 → 7 is greater than or equal

to 18? DO NOT!

20 – 30 8 15 → 15 is greater than or
equal to 18? DO NOT!

30 – 40 6 21 → 21 is greater than or
equal to 18? YES!

40 – 50 3 24

n = 24

As the YES answer came in the fourth college class (30! --- 40), we will say that this will be our
third class Quartile!

Step 4) we will apply the formula of Q3, using data from the Class of Q3, just identified!

46
 Unit 1: Basic Statistics and its Application in ACS

Q3 = 35

Decision

These are values that divide the ordered data set (list) within ten (10) equal parts.

• First Decile (D1) - set value so the data series that 10% of the observations are
smaller than him and 90% are greater.
• According Decile (D2) - set value so the data series that 20% of the observations
are smaller than him and 80% are greater.
• Ninth Decile (D9) - set value so the data series that 90% of the observations are
smaller than him and 10% are greater.

Steps taken to calculate the First Decile:

Determined the n (adding column fi);

Calculate the value of (n / 10) (whether n is even or odd!);

Build the college column;

compare the value of (n / 10) with the college’s values, starting from the first college class (the
top!) and asking the question: “This college is greater than or equal to (n / 10)?” If the answer
is NO, the college spent the next class. When the answer is YES, we will stop and try the
corresponding class! This will be our third class quartile.

Finally, we will apply the formula of Q3, extracting the data Q1 of this class, we just found!
Here is the formula:

Example: For the set below, determine the value of the first decile!

47
 Introduction to Statistics and Probability

Xi fi

0 !--- 10 2

10 !--- 20 5

20 !--- 30 8

30 !--- 40 6

40 !--- 50 3

Step 1) we will find and calculate n (n / 10):

Xi fi

0 !--- 10 2

10 !--- 20 5

20 !--- 30 8

30 !--- 40 6

40 !--- 50 3

n=24

Hence, we find that n = 24 and therefore (n / 10) = 2.4

Steps 2) build the college:

Xi fi fac

0 !--- 10 2 2

10 !--- 20 5 7

20 !--- 30 8 15

30 !--- 40 6 21

40 !--- 50 3 24

n=24

Step 3) we compared the college values with the value of (n / 10), asking the question of
practice, adapted to the first decile:

48
 Unit 1: Basic Statistics and its Application in ACS

Xi fi fac

0 !--- 10 2 2 The second is

greater than or
equal to 2.4? DO
NOT!

10 !--- 20 5 7 7 is the greater

or equal to 2.4?
YES!

20 !--- 30 8 15

30 !--- 40 6 21

40 !--- 50 3 24

n=24

We think, therefore, that the corresponding class (10! --- 20) will be our Class of First Decile!

Step 4) we apply the formula of the First Decile:

To the E: D1 = 10.8

Percentile or centile

Percentiles call or centiles as the ninety-nine values that separate a series of 100 equal parts.
Indicated: P1, P2, ..., P99.

Steps taken to calculate the percentiles or Centis:

To determined the n (adding column fi);

To calculate the value of (Xn / 100) (whether n is even or odd!);

build the college column;

compare the value of (Xn / 100) with the college’s values, starting from the first college class
(the top!) and asking the question: “This college is greater than or equal to (Xn / 100)?” . If the
answer is NO, the college spent the next class. When the answer is YES, we will stop and try
the corresponding class! This is our class X-th centile, or PX class.

Finally, we will apply the formula of PX, PX extracting the data from this class, we just found!
Here is the formula:

MEASURES dispersion or variability

49
 Introduction to Statistics and Probability

Measures of dispersion absolute

a) Total range: It is the only measure of dispersion that has on average the reference point.

When the data are not grouped the total amplitude is adiferença entr the largest and the
smallest observed value: AT = Maximum X - X min.

Example: For the values 40, 45, 48, 62 and 70 will be the total amplitude AT = 70-40 = 30

When data are grouped without class intervals still have AT = maximum X - X min.

Example:

AT=4-0=4

0 2

1 6

2 5

4 3

With class intervals the total amplitude is the difference between the upper limit of the last
class and the lower limit of the first class. Then Range = Lmaximum - L min

Example:

Classes fi

4 |------------- 6 6

6 |------------- 8 2

8 |------------- 10 3

Range = 10-4 = 6

The full range is inconvenient and only consider the two extreme values of the series,
neglecting the set of intermediate values. Makes use of the full range when you want to
determine the amplitude of the temperature in a day, quality control or as a quick calculation
measure without much accuracy.

b) Deviation quartile

Also called semi-interquatílica range and is based on quartiles.

Symbol: Dq and Formula: Dq = (Q3 - Q1) / 2

50
 Unit 1: Basic Statistics and its Application in ACS

Remarks:

1 - The quartile deviation has the advantage the fact that it is an easy measure to calculate
and interpret. Besides, is not affected by extreme, large or small values and is recommended,
therefore, when between the data contained extreme values are not considered representative.

2- The quartile deviation should be used preferably when the measure of central tendency is
the median.

3- It is a measure insensitive to the distribution of smaller items that Q1, between Q1 and Q3
and higher than Q3.

Example: For the values 40, 45, 48, 62 and 70 the quartile deviation is:

Q1 = (45 + 40) / 2 = 42.5 Q3 = (70 + 62) / 2 = 66 Dq = (66 to 42.5) / 2 = 11.75

Standard Deviation

It is the measure most commonly used dispersion because it takes into consideration all the
values of the variable under study. It is an indicator of variability quite stable. The standard
deviation is based on deviations around the mean and its basic formula can be translated as:
the square root of the arithmetic mean of the squares of deviations and is represented by S.

Classes fi

4 |------------- 6 6

6 |------------- 8 2

8 |------------- 10 3

The above formula is used when dealing with a population of non-grouped data.

Example: Calculate the standard deviation of the population represented by - 4, -3, -2, 3, 5

51
 Introduction to Statistics and Probability

We know that n = 5 and 62.8 / 5 = 12.56.

The square root of 12.56 is the standard deviation = 3.54

FMC: Pearson’s Coefficient of Variation

If: CV <15% for low dispersion

If 15% ≤C.V. <30% for average dispersion

Where: RC ≥ 30% for high dispersion

Measures asymmetry:

It is an indicator of the shape of the data distribution

Pearson’s coefficient

52
 Unit 1: Basic Statistics and its Application in ACS

AS = 0 → distribution is symmetrical;

AS> 0 → positive distribution is asymmetric;

AS <0 → distribution is asymmetrical negative.

53
 Introduction to Statistics and Probability

Measures kurtosis:

The degree of flattening of the distribution, is an indicator of the shape of this distribution.

Coefficient of kurtosis

• leptokurtic: when the distribution has a frequency curve rather closed, with the
data strongly concentrated around its center, C <0.263.
• mesokurtic: when data is fairly concentrated around its center, C = 0.263.
• platykurtic: when the distribution has a frequency curve more open with data
weakly concentrated around its center, C> 0.263.

Exercise Application and concentration

1. Consider a sample comprised of discrete data:

9, 8, 5, 4, 5, 6, 2, 2, 4, 3, 4, 7, 9, 5, 6, 7, 1, 4, 7, 2, 4, 6, 3, 5, 7, 9, 5, 1, 4, 8, 2, 9

2. Consider a set of values of measured results. It could be, for example, the age of the
students in the class U of Statistics discipline.

54
 Unit 1: Basic Statistics and its Application in ACS

Age (in months) of the students in the class U - Discipline Statistics

230 234 276 245 345 240 270 310 368 369

334 268 288 336 299 236 239 355 330 247

287 344 300 244 303 248 251 265 246 266

240 320 308 299 312 324 289 320 264 275

252 298 315 255 274 264 263 230 303 281

Answers

1.

55
 Introduction to Statistics and Probability

2.

Conclusion
Practical Exercise:

Group work:

Formative Evaluation

1. The data represent 60 family incomes W Subdivision (data at $ 1,000)

56
 Unit 1: Basic Statistics and its Application in ACS

Is asked:

a) Obtaining descriptive statistics for Microsoft Excel use.

b) Interpret the results

Is asked:

a) Obtaining descriptive statistics for Microsoft Excel use.

b) Interpret the results

57
 Introduction to Statistics and Probability

b)The average income of all the 60 households is $ 7.67797

The standard error = devio ratio between the standard and the square root of n, where n = 60,
is $ 590.

50% of households have incomes of less than $ 8, and the remaining 50%, above this amount
(Note that the median is 8,000)

The fashion equal to $ 3,000, means that the most common income of 60 families group is $
3,000

The dispersion around the mean, as measured by deviation, is $ 4.5331.

The measurement urtose evaluates the degree of flattening of the distribution, and indicates
that the distribution is leptokurtic because the coefficient is negative

It is mildly asymmetric to direct the coefficient is 0.28648

The total amplitude is equal to $ 15,000 (16000-1000)

The lower income is $ 1,000, while the highest is $ 16,000.

The sum of all income reaches 453,000.

58
 Unit 1: Basic Statistics and its Application in ACS

Unit Summary
The teacher will indicate a practical group work field

Unit Assessment
1. Establish which of the following data are discrete and which are continuous:

a) Number of shares sold on the stock exchange

b) Temperatures recorded every half hour in a weather station

c) Length of parts produced by certain machine

d) diameters 1000 fasteners produced by a factory

e) Number of people in the carnival of Brazil

2. What è random trial and in what circumstances should be used?

3. What is probability sampling and when it should be used?

4. A study should be done to determine the annual use in schools. For this, the Ministry of
National Education, has a population of 7000 students spread over 4 levels of education:
3000 are of primary education, secondary education 2000, 1500 high school and 500 are
higher education. The direction of statistical studies of the ministry estimated that the
sample must be at least 700 students to be considered representative.

a) Determine the rate or sampling fraction;

b) Determine, using proportional stratified sampling, the number of

students to be extracted in each stratum of education.

5. To conduct a study on the time spent, in minutes, for 60 elements of a karting club in a
20-lap circuit, there was the time spent by 16 of these elements. The results were as follows:

14.1 13.5 15.0 16.2 17.6 18.7 13.1 15.4

16.6 17.2 14.8 15.9 18.0 16.3 14.9 14.3

a) State:

b) the population;

c) sample.

d) Indicate the study variety and rate it.

e) indicate four values that the statistical variable can take

59
 Introduction to Statistics and Probability

Answers

1. Discrete, Continuous, Continuous, Continuous, Discrete.

2. Each member of the population possesses a certain probability of being selected. Usually
have the same probability. Thus, if N is the population size, the probability of each element
is selected is 1 / N.

3. a) f = n / N = 700/7000 = 0.10 or 10%

b) Primary Education N1 = 3000 students

• A level N2 = 2000 students

• High School N3 = 1500 students
• Higher Education N4 = 500 students

4. Calculating the number of elements to be selected in each stratum of education ni = f *

N

• Elementary School n1 = 0.10 * 3000 = 300

• Secondary education n2 = 0.10 * 2000 = 200
• High school n3 = 0.10 * 1500 = 150
• Higher education n4 = 0.10 * 500 = 50
• The sum of 300 + 200 + 150 + 50 = 700 units in the sample

5 a) 60 elements,

b) 16 elements,

c) time spent in minutes

d) Continuous Quantitative

e) 16.6, 17.2, 14.8 and 15.9

60
 Unit 1: Basic Statistics and its Application in ACS

Readings and other required resources:

Lecture 1

• REIS, Elizabeth. Descriptive Statistics. Lisbon. Issues Syllabus, Lda. Lisbon. 1991
Page 15-46
Lecture 2

• MARTINS, Gilberto de Andrade. General and Applied Statistics. 3rd edition. São
Paulo. Editorial ATLAS S.A; 2005 Pages 19-31

Readings and other optional features:

BRUM PIANA, Clause Fatima. MACHADO, Almeida Amauri. ROLDÃO Selau, Lisiane Priscilla.
Basic Statistics. Pelotas, 2009. Pages 5-12

Internet Resources

www.youtube.com/watch?v=B4L3G30XB7I. 7min Statistics - Basics - Video Lesson

www.portalaction.com.br/content/estatística-básica

It is a statistical software developed for students with easy to use, comprehensive and reliable.
The system was developed under Action R platform, one of the most widely used statistical
systems.

The Action system is a great improvement compared to the statistical software:

2. Allows you to work with Excel in an integrated manner;

3. It is easy to install, creative and covers the main needs of the statistical user;

4. It is becoming more intuitive, easier to use than ever, with a lot of features Action system is
an open and democratic system for the use of statistics:

• This program is free software; you can use it under the terms of the GNU General
Public License;
• No language barrier - is available in Portuguese and English;
• First statistical system that uses the R and Excel platform in an integrated manner,
all to facilitate and expedite their statistical analyzes;

61
 Introduction to Statistics and Probability

www.pt.wikipedia.org/wiki/Estatística

www.alea.pt

ALEA - Local Action of Applied Statistics - is within the scope of Education, Information Society,
the Statistical Information, Training for Citizenship and Literacy Statistics as a contribution
to the development and availability of support instruments to the teaching of Statistics for
students and teachers of Basic and Secondary Education, the main support a web site.

Improving statistical literacy is thus an important condition for, on the one hand, ensure better
provision of a public utility and, on the other hand, foster environments and diverse learning
experiences using new information technologies.

http://www.infoescola.com/estatistica/distribuicao-de-frequencias/

Internet Resources

URL: http: //en.wikipedia.org/wiki/Statistics

It is a democratic feature that is frequently updated, easily accessible through Google, contains
book Probability and Statistics, graduates can find on this site many problems of probability
and statistics content.

http://www.infoescola.com/estatistica/

Site consists of various items of Probability and Statistics, is very practical and easy to access.
This site specifically investigate the theme “coleita methods of data”

http://www.alea.pt

This site presents different statistical concepts with their examples, exercises and intelligent
didactical games.

http://www.pordata.pt/Portugal

www.youtube.com/watch?v=UzBpykJhpyw

Lesson on calculating the mean, median and mode for frequency distribution table for grouped
data

Media: to get the result the average is necessary to add part-mean (x), and then divided by the
sum frequency (f)

M = 3016 = 149.6

Median: to find the result you must use the formula of the precise calculation of the median.

62
 Unit 1: Basic Statistics and its Application in ACS

Unit Summary
We discussed basic statistics and introduction to SPSS. We provided the meaning of statistics,
types of statistics, measures of central tendencies and presentation of data.

Unit Assessment

Check your understanding!

Unit Readings and Other Resources

The readings in this unit are to be found at course level readings and other resources.

63
 Introduction to Statistics and Probability

Unit 3: Linear Regression

Total hours: 40 hours

Introduction to Unit
In prior units, description and the statistical inferences were treated only in terms of a variable.
Thus, when the sample had a business, we considered a variable at a time, for example,
the billing. However when we got a sample of companies, there are several variables that
can be observed in each sampled unit: number of employees, wages, etc. In the first case,
each observation unit is associated with the measurement of a variable X; then, each unit is
associated with the measurements of several variables X, Y, Z, etc.

Unit Goals
Students will be able to:

1. Identify values of a dependent variable (Y) as a function of the independent

variable (X).

2. Describe how changes in X can affect Y.

3. Analyze the simple linear regression model.

Key Terms

• Regression analysis: Allow describe using a mathematical model,

relationship between two variables, starting from n observations of the
same.
• Least squares method: method that allows the adjustment of a straight
line to the observed data
• Dependent variable (RV): Measures the phenomenon that is studied
and we want to explain. Are those whose effects are expected according
to the causes. They are located usually at the end of causal process and
are always set in the case or towards Statistics.
• Independent variable (VI): Those candidates’ variables to explain the
(s) variable (s) dependent (s), the effects of which we want to measure.
Here we must be careful because even finding relationship between
variables that do not necessarily mean causation.

64
 Unit 3: Linear Regression

Learning activities

Activity 1 The simple linear regression model

Total hours: 40 hours

Introduction

Whenever we wish to study particular variable in another function always do a regression

analysis.

We can say that the regression analysis aims to describe, through a mathematical model, the
relationship between two variables, starting from n observations of the same.

Details of the activity

1st Activity: Reading the obligatory references; in order to explain the linear regression model

2nd Activity: Build the scatter diagram by way of example, with the use of Microsoft Excel and
SPSS

3rd Activity: Interpret the scatter plot

The simple linear model regression

In the decision-making process is often necessary to make predictions. At the same time, it
is much easier to make decisions on certain variable when it is possible to establish a link
between this and another variable whose behavior is known.

In order to make predictions about a variable from another variable there must be between the
two a cause and effect relationship, i.e. the variation of a variation can be attributed to another.
The first step in the regression study consists precisely in establishing whether the relationship
variable is not merely accidental.

Once you have established the existence of a possible causal relationship between the
variables, the next step is to study the type of relationship. To do this, should first be making a
scatter plot of the observed data.

Chart Wasting

Graph where each point represents one to observed values (Xi, Yi) corresponding respectively
to the values of the independent and dependent variables. The scatter diagram has a dual
function:

• Helps determine whether there is any relationship between the variables and
• Allows you to identify what is the most appropriate equation to describe this
relationship.

65
 Introduction to Statistics and Probability

The relationships between variables can be of various types: linear, exponential, logarithmic,
power, logistics, etc.

Graphic: different relationships between variables

A: Negative linear relationship

B: Positive linear relationship

C: Ratio of Absence

D: Nonlinear Relationship

The simplest relation is linear type, and you can make many of the linear non-linear
relationships previously identified. The type of linear relationship between two variables can be
described mathematically by the following equation:

Y=a+bX+e Simple linear regression model

Where:

• Y is the explained or dependent variable;

• X is the explanatory variable or independent;
• And a residual variable type that includes other Y explanatory factors not included
in X and even measurement errors;
• A and b are constants: is the intercept of the straight line with the vertical axis e b
o slope of the line.

66
 Unit 3: Linear Regression

Example:

Suppose, for example, that an economist studies the relationship between unit labor cost and
the price index of producer in order to make predictions about the last variable from known
values of the former. For such data is available from 1984 to 1990:

Growth of unit labor costs and producer prices

year Growth of unit cost Price growth

labor (%) the producer (%)

1984 7,8 10,8

1985 5,7 4,4

1986 6,1 6,5

1987 7,7 7,8

1988 11,2 11,1

1989 11,2 13,5

1990 8,3 9,2

Once the economist is interested in predicting changes in the production price, this is defined
as the dependent variable and will be called Y. prediction will be based on the independent
variable called X and that in this concrete example is the growth of unit labor cost. In SPSS.

1 From the menu bar choose:

Scatter Graphs →....

Select simple scatter

67
 Introduction to Statistics and Probability

• Select the dependent variable for the Y axis.

• Select the independent variable to the axis X.

68
 Unit 3: Linear Regression

·To identify the points, use the Data mode ID option in the Chart Editor window

69
 Introduction to Statistics and Probability

To view the line:

The relationship between the two variables is linear, positive type, because the higher the
growth of the labor cost, the greater the increase in the price of the manufacturer, i.e., where
the two variables vary in the same direction.

70
 Unit 3: Linear Regression

Conclusion

Linear regression permit find the line that best represents the relationship between two
variables.

Example: Simple Linear Regression - Excel software Use

We have a set of 5,000 observations of the variables X and Y has been registered and the
equation of the straight line was obtained

See the column “coefficients” to make sure the values of the parameters a and b of the line.

Formative Evaluation

Preparation of research papers, individual, based on the regression analysis applied to different
areas or professions (medicine, agriculture, economics, etc ...), so that the students feel
encouraged to use the resources available on the Internet. All materials produced in this area
will be an integral part of the individual electronic portfolio forming.

Exercise Application and concentration

1.An interviewer’s administrator wants to develop a model to predict the number of interviews
in a given day. He believes the interviewer’s experience (measured in weeks worked) is
determining the number of interviews. A sample of 10 interviewers gave the following data:

Week 15 41 58 18 37 52 28 24 45 33
experience

Number of 4 9 12 6 8 10 6 5 10 7
interviews

71
 Introduction to Statistics and Probability

Naming Y = number of interviews and X = Weeks experience, we can build the scatter diagram
in SPSS.

2. A study was conducted in order to investigate whether there is any relationship between
agricultural production and energy consumption and ultimately, it is possible to predict
agricultural production from energy consumption.

Considering the data for 9 years.

Year Agricultural Contents

production index of energy
(1997 = 100) consumption
(1997 = 100)

1997 100 100

1999 104 112

2002 111 121

2004 127 131

2006 133 1374

2008 139 162

2010 144 185

2012 144 193

2013 173 219

a) Which of the variables should be considered explanatory?

b) Draw a scatter plot and describe the type of relationship between the variables

2. The table below shows the frequency of the average pulse in different age periods:

72
 Unit 3: Linear Regression

Age pulse

2 112

4 104

6 100

8 92

10 88

12 86

14 84

16 80

a) Find the linear regression equation

4. A sample of plants has led to an industry:

Total Cost Y Production X

80 12

44 4

51 6

70 11

61 8

a) Find the linear regression equation.

b) What are the economic meanings of “a” and “b”?

c) Find the coefficient of determination (or explanation).

d) Test the existence of regression to a 5% significance level.

e) Determine a prediction interval (90%) for a given X Y mean = 10

73
 Introduction to Statistics and Probability

Answers

1.

3. a) Index do consumption of energy

b)

74
Unit 3: Linear Regression

75
 Introduction to Statistics and Probability

4.

76
 Unit 3: Linear Regression

Activity 2 Least squares

Introduction

The least squares method enables the adjustment of a straight Linea observed data so that
minimizes the sum of squared distances between the observed values and the fitted line, these
measures the vertical distance and exactly corresponding to the differences between the
observed values Y and the adjusted values of

1st Activity: Construct scatter diagram with an example

When adjusting one regression line to the observed data, the linear relationship between the
two variables becomes great because they cancel all purposes residual variable. Adjusted
straight, will then have the mathematical form:

Regression line

Deviation between the observed values and the adjusted values

More specifically, for a given value of the independent variable Xi will have two values for Y: an
observed value Yi and another which is given by the set straight Yai. The difference between
the two, how easily it shows exactly matches the random effect will residue:

Being and ,

Then .

77
 Introduction to Statistics and Probability

By applying the method of least squares is intended to fit a straight line which minimizes the
square sum of i waste, i.e., find values for the constants a and b that make this minimum sum:

The least squares method allows us to find a regression line, whose coefficients are given by:

• Intercept of the regression line

• Slope of the regression line

And that, for any line, has the advantage of being an optimal solution in that makes minimum
distances between observed values for Y and the straight set.

Example

For example presented in the previous activity, we intend to fit a least squares straight to the
observed data in order to predict changes in producer prices. First you need to calculate be
for it must be known the following sums: and. Once be calculated by knowing the average of X
and Y, calculating the immediate.

Year Yi Xi XiYi Xi2

1984 10,8 7,8 84,24 60,84

1985 4,4 5,7 25,08 32,49

1986 6,5 6,1 39,65 37,21

1987 7,8 7,7 60,06 59,29

1988 11,1 11,2 124,32 125,44

1989 13,5 11 148,5 121

1990 9,2 8,3 76,36 68,89

Number of observations n = 7

78
 Unit 3: Linear Regression

Graphically, this regression line has the following configuration

Conclusion
How to Interpret the coefficients a and b?

In a regression line, the calculated value for the coefficient gives us exactly the point of
intersection of the line with the axis of the dependent variable. This value can be positive
or negative and indicates the very general position of the regression line. Its simplest
interpretation is as follows:

It corresponds to the value of the dependent variable Y when the effect of the independent
variable X is null. For this reason the coefficient is referred to as constant.The b coefficient of
the regression line exactly matches the slope of this line

and therefore gives us know setting this straight: if it is positive the line will have
a positive slope and vice versa: the greater its value, the stronger this slope. The coefficient b
represents the expected variation of the dependent variable Y for each unit of the independent
variable X. For example, if Y represents sales of a product X, and the advertising expenditure
for the same product, if b = 8.0, this means that by $ 1,000 of increased spending on
advertising, sales will increase $ 8,000.

Exercise

1.Consider the data:

X 0 1 2 3

Y 1 2 4 5

79
 Introduction to Statistics and Probability

a) Calculate xiyi

b) Calculate XiXi

c) Calculate X

d) Find b

e) Find a

2. The data below refer to the volume of rainfall (mm) and the volume of milk production type
C (million liters) in a certain region of the country.

Set the data using a linear model

Granted, in 1980, a 24 mm rainfall, which should be expected volume of the milk type C?

Year Milk production Rainfall Index

(1,000,000 l) (mm)

1970 26 23

1917 25 21

1972 31 28

1973 29 27

1974 27 23

1975 31 28

1976 32 27

1977 28 22

1978 30 26

1979 30 25

3.Consider the data

X -5 -3 0 1 3 5

Y 0,8 1,1 2,5 3,1 5,0 4,7 6,2

a) Build the scatter diagram.

b) Find the line of least squares

4.For a company to remain competitive, spending on research and development (R & D) are
essential. To determine the optimal level of spending on R & D and its effect on the value of
the company, was applied simple linear regression analysis, where:

80
 Unit 3: Linear Regression

Y = ratio of prices and earnings

X = ratio of expenditure on R & D and sales

The following data of 20 companies used in the study are:

Empresas y x Empresas y x

1 5,6 0,003 11 8,4 0,058

2 7,2 0,004 12 11,1 0,058

3 8,1 0,009 13 11,1 0,067

4 9,9 0,021 14 13,2 0,080

5 6,0 0,023 15 13,4 0,080

6 8,2 0,030 16 11,5 0,083

7 6,3 0,035 17 9,8 0,091

8 10,0 0,037 18 16,1 0,092

9 8,5 0,044 19 7,0 0,064

10 13,2 0.051 20 5,9 0,028

a) Build the scatter plot

b) Adjust the line of least squares

c) Using the equation obtained to predict the value of y when x = $ 0.070

solution

1.a) 15

b) 10

c) 1

d) 1.50

e) 0.70

81
 Introduction to Statistics and Probability

2.

Y X X2 XY

26 23 529 598

25 21 441 525

31 28 784 868

29 27 729 783

27 23 529 621

31 28 784 868

32 27 729 864

28 22 484 616

30 26 676 780

30 25 625 750

Sy = 289 Sx = 250 Sx2 =6.310 Sxy = 7.273

Determine the value of the parameter b

II - Determining the value of the parameter

82
 Unit 3: Linear Regression

a = 289 - 0,8. 250 = 8,9

10 10

III - Equation of Straight Adjusted

y = a + bx

y = 8,9 +0,8x

b) setting x = 24 mm have y = 8.9 + 28.1 = 0,8x24.

According co the model, we can expect 28.1 million liters produced for a 24 mm rainfall

3. a)

b)

83
 Introduction to Statistics and Probability

4.

84
 Unit 3: Linear Regression

Unit Summary
Linear regression is a statistical tool used to predict future values from past values. The linear
regression trend line uses squares method to draw a straight line through prices in order to
minimize the distance between them and the resulting trend line.

Linear Regression is a statistical issue of great importance and applicability, not only the
disciplines and related professions such as mathematics, engineering, statistics among others,
but also realize its application in various areas such as medicine, pharmacology and even
in music . Studying this topic will help the individual to improve their statistical perception
providing you a complete logical reasoning.

Unit Assessment

• Construction and interpretation of the scatter gram, the use of Excel and SPSS
• Exercises on the method of least squares
• Graphical straight adjustment Representation
• Application of mathematical model

Assessment tools

• Construction of PORTFOLIOS
• Resolution of exercises
• Reporting

Evaluation criteria

Number and type of coefficient

evaluation

1 Group work 10%

2 individual work 15%

4 Solving exercises 20%

3 Research work 15%

1 Case Study 15%

1 Final exam 25%

85
 Introduction to Statistics and Probability

Evaluation exercise

Considering the data in the following table:

Provided heat
units

11,878 179476

13,087 190724

11,623 173965

13,474 196530

11,584 172064

10,949 162246

14,52 212716

4,056 59639

14,344 211407

13,316 194961

15,852 233603

13,26 194932

14,69 213024

100000

By using the SPSS

a) Determine the equation of the regression line

b) Build the scatter diagram.

86
 Unit 3: Linear Regression

Solution

a) Y= -0, 869 + 7,20 x 10-5 x

Lecture and other resources

• www.youtube.com/watch?v=x42skwrbiekhttps://www.youtube.com/
watch?v=L_glrTzMd7c
• https://www.google.pt/#q=Regres%C3%A3o+Linear+simplespt.slideshare.
net/monica_lima/regresso-linear-simpleshttp://www.youtube.com/
watch?v=L_glrTzMd7c.

87
 Introduction to Statistics and Probability

Unit 4: Applications of Probability

and Statistics in ACS
Unit Introduction
Statistics may have little to offer the search architectures in a data mining search,

but a great deal to offer in evaluating hypotheses in the search, in evaluating the results of the
search, and in applying the results in computing. Probability and statistics have a widespread
applications in the area of applied science e.g artificial intelligence, Virtual reality.

Unit Objectives
Upon completion of this unit you should be able to:

1. Apply probability and statistics in machine learning.

2. Solve problems in ACS using Bayesian methods

3. Define Supervised and unsupervised learning.

4. Identify the different types of classification and give examples.

5. Define clustering and gives examples.

Key Terms
Machine Learning: It builds statistical models of data in
order to recognize complex patterns and to make decisions
based on observed data . Examples include: classification
(recognition of faces or handwritings, predictions (stock
market, election), data mining, etc.

Labeled Data: there is a specially designated attribute

and the aim is to use the data given to predict the value
of that attribute for instances that have not yet been seen.
Data of this kind is called labeled.

Unlabeled Data: Data that does not have any specially

designated attribute is called unlabelled.

Supervised learning: Data mining using labeled data is

known as supervised learning.

Unsupervised learning: Data mining of unlabelled data is

known as unsupervised learning.

88
 Unit 4: Applications of Probability and Statistics in ACS

Classification: is a task that occurs very frequently in

everyday life. Essentially it involves dividing up objects
so that each is assigned to one of a number of mutually
exhaustive and exclusive categories known as classes.

Nearest Neighbor Matching: This method relies on

identifying (say) the five examples that are ‘closest’ in
some sense to an unclassified one.

Training Data Set: The training set constitutes the

results of a sample of trials that we can use to predict the
classification of other (unclassified) instances.

Instance: An instance comprises the values of a number of

attributes and the corresponding classification.

Classification Tree: One way of generating classification

rules is via an intermediate tree-like structure called a
classification tree or a decision tree.

Neural Network: This is a complex modeling technique

based on a model of a human neuron.

Association Rules: A training set is use to find any

relationship that exists amongst the values of variables,
generally in the form of rules known as association rules.

Clustering: Clustering algorithms examine data to find

groups of items that are similar. For example, an insurance
company might group customers according to income,
age, types of policy purchased or prior claims experience.

Learning Activities

Activity 1 Naive Bayesian and k-means neighbor

Introduction

Classification is a task that occurs very frequently in everyday life. Essentially it involves dividing
up objects so that each is assigned to one of a number of mutually exhaustive and exclusive
categories known as classes. The term ‘mutually exhaustive and exclusive’ simply means that
each object must be assigned to precisely one class, i.e. never to more than one and never to
no class at all. For example, a hospital may want to classify medical patients into those who are
at high, medium or low risk of acquiring a certain illness, an opinion polling company may wish
to classify people interviewed into those who are likely to vote for each of a number of political
parties or are undecided, or we may wish to classify a student project as distinction, merit, pass
or fail.

89
 Introduction to Statistics and Probability

Classification for supervised learning can be done in two ways: Naive Baye’s and K-means
neighbor

Activity Details

Lesson 1: Naïve Baye’s classifier

The Naive Bayes algorithm gives us a way of combining the prior probability and conditional
probabilities in a single formula, which we can use to calculate the probability of each of the
possible classifications in turn.

For example, a fruit may be considered to be an apple if it is red, round, and about 3” in
diameter. A naive Baye’s classifier considers each of these features to contribute independently
to the probability that this fruit is an apple, regardless of any possible correlations between the
color, roundness and diameter features.

Example 1:

Given a set of k mutually exclusive and exhaustive classifications

which have prior probabilities , respectively, and n

attributes for a given instance have values respectively, the

posterior probability of class occurring for the specified instance can be shown to be

proportional to Making

the assumption that the attributes are independent, the

value of this expression can be calculated using the product

We calculate this product for each value of i from 1 to k and choose the classification that has
the largest value.

Example 2:

When dealing with continuous data, a typical assumption is that the continuous values
associated with each class are distributed according to a Gaussian distribution. For example,
suppose the training data contain a continuous attribute . We first segment the data by the
class, and then compute the mean and variance of in each class. Let be the mean of the
value in associated with class c, and let be the variance of the values in associated
with class c. Then, the probability distribution of some value given a class,
can be computed by into the equation for a Normal Distribution parameterized by and
. That is,

90
 Unit 4: Applications of Probability and Statistics in ACS

For some types of probability models, naive Bayes classifiers can be trained very efficiently in a
supervised learning setting.

0000000000000000000000000000000000000000000000000000000000000000000000000
00000000000000000000000000000000000000000000000000000000000000000000000000
0000000000000000000000000000000000000000000000000000000000000000000000000000
00

Exercise: Classify whether a given person is a male or a female based on the measured
features. The features include height, weight, and foot size.

Solution:

Using the training set below

sex height (feet) weight (lbs) foot size(inches)

male 6 180 12

male 5.92 (5’11”) 190 11

male 5.58 (5’7”) 170 12

male 5.92 (5’11”) 165 10

female 5 100 6

female 5.5 (5’6”) 150 8

female 5.42 (5’5”) 130 7

Assuming that we have equi-probable classes so P(male)= P(female) = 0.5. This prior
probability distribution might be based on our knowledge on frequency in the training set.

Testing

Below is a sample to be classified as a male or female.

sex height (feet) weight (lbs) foot size(inches)

sample 6 130 8

We wish to determine which posterior is greater, male or female. For the classification as male
the posterior probability is given by

For the classification as female the posterior probability is given by

The evidence (also termed normalizing constant) may be calculated:

91
 Introduction to Statistics and Probability

We now determine the probability distribution for the sex of the sample.

where and are the parameters of normal

distribution which have been previously determined from the training set. Note that a value
greater than 1 is OK here – it is a probability density rather than a probability, because height is
a continuous variable.

Since posterior numerator is greater in the female case, we predict the sample is female.

Lesson 2: k-nearest neighbor (k-NN)

In practice there are likely to be many more instances in the training set but the same principle
applies. It is usual to base the classification on those of the k nearest neighbors (where k is a
small integer such as 3 or 5), not just the nearest one. The method is then known as k-Nearest
Neighbor or just k-NN classification.

Basic k-Nearest Neighbor Classification Algorithm

• Find the k training instances that are closest to the unseen instance.
• Take the most commonly occurring classification for these k instances.

Supposing we have a training set with just two instances such as the following:

92
 Unit 4: Applications of Probability and Statistics in ACS

a b c d e f class

yes no no 6.4 8.3 low negative

yes yes yes 18.2 4.7 high positive

There are six attribute values, followed by a classification (positive or negative). We are then
given a third instance

yes no no 6.6 8 low ???

What should its classification be?

Even without knowing what the six attributes represent, it seems intuitively obvious that the
unseen instance is nearer to the first instance than to the second. In the absence of any other
information, we could reasonably predict its classification using that of the first instance, i.e. as
‘negative’.

We can illustrate k-NN classification diagrammatically when the dimension (i.e. the number of
attributes) is small. The following example illustrates the case where the dimension is just 2.

93
 Introduction to Statistics and Probability

Attribute 1 Attribute 2 Class

0.8 6.3 −

1.4 8.1 −

2.1 7.4 −

2.6 14.3 +

6.8 12.6 −

8.8 9.8 +

9.2 11.6 −

10.8 9.6 +

11.8 9.9 +

12.4 6.5 +

12.8 1.1 −

14 19.9 −

14.2 18.5 −

15.6 17.4 −

15.8 12.2 −

16.6 6.7 +

17.4 4.5 +

18.2 6.9 +

19 3.4 −

19.6 11.1 +

Table1

Table 1 show a training set with 20 instances, each giving the values of two attributes and
an associated classification. How can we estimate the classification for an ‘unseen’ instance
where the first and second attributes are 9.1 and 11.0, respectively? For this small number
of attributes we can represent the training set as 20 points on a two-dimensional graph with
values of the first and second attributes measured along the horizontal and vertical axes,
respectively. Each point is labeled with a + or − symbol to indicate that the classification is
positive or negative, respectively. The result is shown in Figure 2.

94
 Unit 4: Applications of Probability and Statistics in ACS

Using the training dataset in the table above, the figure below is obtained.

A circle has been added to enclose the five nearest neighbors of the unseen instance, which is
shown as a small circle close to the centre of the larger one.

The five nearest neighbors are labeled with three + signs and two − signs, so a basic 5-NN
classifier would classify the unseen instance as ‘positive’ by a form of majority voting. There are
other possibilities, for example the ‘votes of each of the k nearest neighbors can be weighted,
so that the classifications of closer neighbors are given greater weight than the classifications
of more distant ones.

We can represent two points in two dimensions (‘in two-dimensional space’ is the usual term)
as

and and visualize them as points in a plane.

When there are three attributes we can represent the points by and

and think of them as points in a room with three axes at right angles. As the
number of dimensions (attributes) increases it rapidly becomes impossible to visualize them, at
least for anyone who is not a physicist (and most of those who are).

When there are n attributes, we can represent the instances by the points
and in ‘n-dimensional space’.

95
 Introduction to Statistics and Probability

Conclusion

The Naive Bayes approach is a very popular one, which often works well. However it has a
number of potential problems, the most obvious one being that it relies on all attributes being
categorical. In practice, many datasets have a combination of categorical and continuous
attributes, or even only continuous attributes. This problem can be overcome by converting
the continuous attributes to categorical ones, using a different method.

Group Activity

In a group of five students, answer the following questions. It should be submitted after one
week. The total mark is out of 50.

1. Using the Naiıve Bayes classification algorithm with the train dataset in the table below,
calculate the most likely classification for the following unseen instances.

Testing Set:

weekday summer high heavy ????

sunday summer normal slight ????

96
 Unit 4: Applications of Probability and Statistics in ACS

Training Data set:

Day Season Wind Rain Class

weekday Spring none none On time

weekday Winter none slight On time

weekday Winter none slight On time

weekday Winter high heavy late

Saturday Summer normal none On time

weekday Autumn normal none Very late

holiday Summer high slight On time

Sunday Summer normal none On time

weekday Winter high heavy Very late

weekday Summer none slight On time

Saturday Spring high heavy cancelled

weekday Summer high slight On time

Saturday Winter normal none late

weekday summer high none On time

weekday Winter normal heavy Very late

Saturday Autumn high slight On time

weekday Autumn none heavy On time

holiday Spring normal slight On time

weekday Spring normal none On time

weekday Spring normal slight On time

Using the training set shown in Table 1 and the Euclidean distance measure, calculate the
5-nearest neighbors of the instance with first and second attributes 9.1 and 11.0, respectively.

Activity 2 Decision Tree for classification

Introduction

In this activity students will learn how to represent data graphically and derive information from
it.

97
 Introduction to Statistics and Probability

Activity Details

Decision tree learning is the construction of a decision tree from class-labeled training tuples.
A decision tree is a flow-chart-like structure, where each internal (non-leaf) node denotes a test
on an attribute, each branch represents the outcome of a test, and each leaf (or terminal) node
holds a class label. The topmost node in a tree is the root node.

Example

We have a dataset in the form of a table containing students’ grades on five subjects
(the values of attributes SoftEng, ARIN, HCI, CSA and Project) and their overall degree
classifications. The row of dots indicates that a number of rows have been omitted in the
interests of simplicity. We want to find some way of predicting the classification for other
students given only their grade ‘profiles’.

Soft Eng ARIN HCI CSA Project Class

A B A B B Second

A B B B B Second

B A A B A Second

A A A A B First

A A B B A First

B A A B B Second

…………. …………. …………. …………. …………. ………….

B A A B First

One way of generating classification rules is via an intermediate tree-like structure called a
classification tree or a decision tree.

The figure below shows a possible decision tree corresponding to the degree classification
data.

98
 Unit 4: Applications of Probability and Statistics in ACS

Figure 3: Decision Tree for Degree Classification Data

The decision tree can be used for classification purpose. Refer to Principles of Data Mining –
Max Bramer (Page 42 – 47).

For reading purpose, refer to

• http://ocw.mit.edu/courses/sloan-school-of-management/15-097-
• prediction-machine-learning-and-statistics-spring-
• 2012/lecturenotes/MIT15_097S12_lec08.pdf

Conclusion

In this activity, the use of tree diagram in computer science is illustrated. It is a very useful way
in organizing the data set presented to them.

Group Activity

This is an activity to be done by a group of five (5). Its purpose is to help the students better
understand and apply the concepts learned previously.

A golfer decides whether or not to play each day on the basis of the weather. The table
below shows the results of two weeks (14 days) of observations of weather conditions and the
decision on whether or not to play.

99
 Introduction to Statistics and Probability

Outlook Temp (◦F) Humidity Windy Class

Sunny 75 70 1 play

Sunny 80 90 1 Don’t play

Sunny 85 85 0 Don’t play

Sunny 72 95 0 Don’t play

Sunny 69 70 0 play

Overcast 72 90 1 play

Overcast 83 78 0 play

Overcast 64 65 1 play

Overcast 81 75 0 play

Rain 71 80 1 Don’t play

Rain 65 70 1 Don’t play

Rain 75 80 0 play

Rain 68 80 0 play

Rain 70 96 0 play

Classes

play, don’t play

Outlook

sunny, overcast, rain

Temperature

numerical value

Humidity

numerical value

Windy

true, false

With the help of a decision tree, assuming the golfer is acting consistently, what are the
rules that determine the decision whether or not to play each day? If tomorrow the values of
Outlook, Temperature, Humidity and Windy were sunny, 74◦F, 77% and false respectively, what
would the decision be?

100
 Unit 4: Applications of Probability and Statistics in ACS

Activity 3 Clustering

Introduction

Clustering is concerned with grouping together objects that are similar to each other and
dissimilar to the objects belonging to other clusters.

In many fields there are obvious benefits to be had from grouping together similar objects.
There are various ways of clustering:

1. Euclidean distance

2. k-means clustering

Activity Details

Lesson 1: Euclidean Distance

Typically, the basic data used to form clusters is a table of measurements on several variables
where each column represents a variable and a row represents an object often referred to in
statistics as a case. Thus the set of rows are to be grouped so that similar cases are in the same
group. The number of groups may be specified or has to be determined from the data.

A popular distance measure based on variables that take on continuous values is to standardize
the values by dividing by the standard deviation (sometimes other measures such as range
are used) and then to compute the distance between objects using the Euclidean metric. The
Euclidean distance between two cases, i and j with variable

values and is defined by:

Example: Public Utilities Data (corporate data on 22 US public utilities)

101
 Introduction to Statistics and Probability

No. Company X1 X2 X3 X4 X5 X6 X7 X8

1 Arizona Public Service 1.06 9.2 151 54.4 1.6 9077 0 0.628

2 Boston Edison Company 0.89 10.3 202 57.9 2.2 5088 25.3 1.555

3 Central Louisiana Electric 1.43 15.4 113 53 3.4 9212 0 1.058

Co.

4 Commonwealth Edison Co. 1.02 11.2 168 56 0.3 6423 34.3 0.7

5 Consolidated Edison Co. 1.49 8.8 1.92 51.2 1 3300 15.6 2.044
(NY)

6 Florida Power and Light 1.32 13.5 111 60 -2.2 11127 22.5 1.241

7 Hawaiian Electric Co. 1.22 12.2 175 67.6 2.2 7642 0 1.652

8 Idaho Power Co. 1.1 9.2 245 57 3.3 13082 0 0.309

9 Kentucky Utilities Co. 1.34 13 168 60.4 7.2 8406 0 0.862

10 Madison Gas & Electric Co. 1.12 12.4 197 53 2.7 6455 39.2 0.623

11 Nevada Power Co. 0.75 7.5 173 51.5 6.5 17441 0 0.768

12 New England Electric Co. 1.13 10.9 178 62 3.7 6154 0 1.897

13 Northern States Power Co. 1.15 12.7 199 53.7 6.4 7179 50.2 0.527

14 Oklahoma Gas and Electric 1.09 12 96 49.8 1.4 9673 0 0.588

Co.

15 Pacific Gas & Electric Co. 0.96 7.6 164 62.2 -0.1 6468 0.9 1.4

16 Puget Sound Power & 1.16 9.9 252 56 9.2 15991 0 0.62
Light Co.

17 San Diego Gas & Electric 0.76 6.4 136 61.9 9 5714 8.3 1.92
Co.

18 The Southern Co. 1.05 12.6 150 56.7 2.7 10140 0 1.108

19 Texas Utilities Co. 1.16 11.7 104 54 -2.1 13507 0 0.636

20 Wisconsin Electric Power 1.2 11.8 148 59.9 3.5 7297 41.1 0.702
Co.

21 United Illuminating Co. 1.04 8.6 204 61 3.5 6650 0 2.116

22 Virginia Electric & Power 1.07 9.3 1784 54.3 5.9 10093 26.6 1.306
Co.

102
 Unit 4: Applications of Probability and Statistics in ACS

X1: Fixed-charge covering

ratio (income/debt)

X2: Rate of return on capital

X3: Cost per KW capacity in

place

X4: Annual Load Factor

X5: Peak KWH demand

growth from 1974 to
1975

X6: Sales (KWH use per year)

X7: Percent Nuclear

X8: Total fuel costs (cents per

KWH)

We are interested in forming groups of similar utilities. The objects to be clustered are the
utilities. There are 8 measurements on each utility described in the table below. An example
where clustering would be useful is a study to predict the cost impact of deregulation. To do
the requisite analysis economists would need to build a detailed cost model of the various
utilities. The objects to be clustered are the utilities and there are 8 measurements on each
utility.

The idea behind this set of techniques is to start with each cluster comprising of exactly one
object and then progressively agglomerating (combining) the two nearest clusters until there is
just one cluster left consisting of all the objects.

Nearness of clusters is based on a measure of distance between clusters. All agglomerative

methods require as input a distance measure between all the objects that are to be clustered.
This measure of distance between objects is mapped into a metric for the distance between
clusters (sets of objects) metrics for the distance between two clusters.

The results of the distance are calculated using SPSS and are shown in the table 1 below.

2. Nearest neighbor clustering

Here the distance between two clusters is defined as the distance between the nearest pair
of objects with one object in the pair belonging to a distinct cluster. If cluster A is the set
of objects and cluster B is the single linkage distance between A
and B is . This method has a tendency to cluster together at an
early stage objects that are distant from each other in the same cluster because of a chain of
intermediate objects in the same cluster. Such clusters have elongated sausage-like shapes
when visualized as objects in space.

103
 Introduction to Statistics and Probability

3. Group Average (also called average linkage).

Here the distance between two clusters is defined as the average distance between all
possible pairs of objects with one object in each pair belonging to a distinct cluster. If
cluster A is the set of objects and cluster B is , the single linkage
distance between A and B is the sum being taken over

The nearest neighbor clusters for the utilities are displayed in Figure 1 below in a useful
graphic format called a Dendogram. For any given number of clusters we can determine the
cases in the clusters by sliding a vertical line from left to right until the number of horizontal
intersections of the vertical line equals the desired number of clusters. For example, if we
wanted to form 6 clusters we would find that the clusters are: {1, 18, 14, 19, 9, 10, 13, 4, 20,
2, 12, 21, 7, 15, 22, 6}; {3}; {8, 16}; {17}; {11}; and {5}. Notice that if we wanted 5 clusters they
would be the same as for six with the exception that the first two clusters above would be
merged into one cluster. In general all hierarchical methods have clusters that are nested within
each other as we decrease the number of clusters we desire.

104
 Unit 4: Applications of Probability and Statistics in ACS

Table1: Distances based on standardized variable values.

For the average linkage, SPSS is used to construct a hierarchical cluster. It is illustrated in Figure
2.

Figure2: Average Linkage

Average Linkage (Between Groups)

Agglomeration Schedule

105
 Introduction to Statistics and Probability

Agglomeration Schedule

Stage Cluster Combined Coefficients Stage Cluster First Next

Appears Stage

Cluster 1 Cluster 2 Cluster 1 Cluster 2

1 4 10 1905.226 0 0 2

2 4 15 3024.826 1 0 5

3 13 20 16655.503 0 0 7

4 1 3 19712.962 0 0 11

5 4 21 42826.06 2 0 6

6 4 12 127946.985 5 0 12

7 7 13 169591.332 0 3 12

8 8 19 200550.521 0 0 19

9 14 18 221054.932 0 0 13

10 2 17 396598.6 0 0 15

11 1 9 551673.97 4 0 13

12 4 7 954407.042 6 7 15

13 1 14 1196900.67 11 9 16

14 11 16 2108774.49 0 0 19

15 2 4 2249966.5 10 12 18

16 1 6 3674274.889 13 0 17

17 1 22 3715771.471 16 0 20

18 2 5 1.08E+7 15 0 20

19 8 11 1.23E+7 8 14 21

20 1 2 1.43E+7 17 18 21

21 1 8 6.27E+7 20 19 0

106
 Unit 4: Applications of Probability and Statistics in ACS

Conclusion

This unit illustrates how important probability and statistics is important to computer sciences.
There are various applications statistics into it specially in analyzing the results obtained in
computing. Few examples have been used and the results are presented using SPSS

Unit Summary
In this unit, students were presented with some basic applications of probability and statistics
in applied computer science. It illustrates also the idea that SPSS could be used to solve most
of the statistical problems in computer science. More applications of could be find in the
reference text. Throughout the unit, students did learn about classification, nearest neighbor,
Naïve Baye’s, etc.

107
 Introduction to Statistics and Probability

Unit Assessment
Check your understanding!

Assignment

Instructions

After reading the case use the data provided to answer the questions that follow

Case: German Credit The German Credit data set (available at ftp.ics.uci.edu/pub/machine-
learning-databases/statlog/) contains observations on 30 variables for 1000 past applicants
for credit. Each applicant was rated as “good credit” (700 cases) or “bad credit” (300 cases).
New applicants for credit can also be evaluated on these 30 “predictor” variables. We want to
develop a credit scoring rule that can be used to determine if a new applicant is a good credit
risk or a bad credit risk, based on values for one or more of the predictor variables. All the
variables are explained in Table 1.1. (Note: The original data set had a number of categorical
variables, some of which have been transformed into a series of binary variables so that they
can be appropriately handled by SPSS. Several ordered categorical variables have been left
as is; to be treated by SPSS as numerical. The data has been organized in the spreadsheet
German CreditI.xls)

The data set and the problem (2) can be obtained from this link: http://ocw.mit.edu/courses/
sloan-school-of-management/15-062-data-mining-spring-2003/assignments/

Table 1.1 Variables for the German Credit data.

Codelist

Var. # Variable Name Description Variable Code Description

Type

1.   OBS# Observation No. Categorical

2.   CHK_ACCT Checking Categorical 0 : < 0 DM

account status

1: 0 < ...< 200 DM

2 : => 200 DM

3: no checking
account

3.   DURATION Duration of credit Numerical

in months

108
 Unit 4: Applications of Probability and Statistics in ACS

4.   HISTORY Credit history Categorical 0: no credits taken

1: all credits at this

bank paid back
duly

2: existing credits
paid back duly till
now

3: delay in paying
off in the past

4: critical account

5.   NEW_CAR Purpose of credit Binary car (new) 0: No,

1: Yes

6 USED_CAR Purpose of credit Binary car (used) 0: No,

1: Yes

7 FURNITURE Purpose of credit Binary furniture/

equipment 0:
No, 1: Yes

8 RADIO/TV Purpose of credit Binary radio/television

0: No, 1: Yes

9 EDUCATION Purpose of credit Binary education 0: No,

1: Yes

10 RETRAINING Purpose of credit Binary retraining 0: No,

1: Yes

11 AMOUNT Credit amount Numerical

12 SAV_ACCT Average balance Categorical 0 : < 100 DM

in savings
account

1 : 100<= ... <

500 DM

2 : 500<= ... <

1000 DM

3 : =>1000 DM

4 : unknown/ no
savings account

109
 Introduction to Statistics and Probability

13 EMPLOYMENT Present Categorical 0 : unemployed

employment
since

1: < 1 year

2 : 1 <= ... < 4

years

3 : 4 <=... < 7
years

4 : >= 7 years

14 INSTALL_RATE Installment Numerical

rate as % of
disposable
income

15 MALE_DIV Applicant is male Binary 0: No, 1: Yes

and divorced

16 MALE_SINGLE Applicant is male Binary 0: No, 1: Yes

and single

17 MALE_MAR_WID Applicant is male Binary 0: No, 1: Yes

and married or a
widower

18 CO-APPLICANT Application has a Binary 0: No, 1: Yes

co-applicant

19 GUARANTOR Applicant has a Binary 0: No, 1: Yes

guarantor

20 PRESENT_ Present resident Categorical 0: <= 1 year

RESIDENT since - years

1<…<=2 years

2<…<=3 years

3:>4years

21 REAL_ESTATE Applicant owns Binary 0: No, 1: Yes

real estate

22 PROP_UNKN_ Applicant owns Binary 0: No, 1: Yes

NONE no property (or
unknown)

23 AGE Age in years Numerical

110
 Unit 4: Applications of Probability and Statistics in ACS

24 OTHER_INSTALL Applicant has Binary 0: No, 1: Yes

other installment
plan credit

25 RENT Applicant rents Binary 0: No, 1: Yes

26 OWN_RES Applicant owns Binary 0: No, 1: Yes

residence

27 NUM_CREDITS Number of Numerical

existing credits at
this bank

28 JOB Nature of job Categorical 0 : unemployed/

unskilled
- non-resident

1 : unskilled
- resident

2 : skilled
employee / official

3 : management/
self-employed/
highly qualified
employee/ officer

29 NUM_ Number of Numerical

DEPENDENTS people for whom
liable to provide
maintenance

30 TELEPHONE Applicant has Binary 0: No, 1: Yes

phone in his or
her name

31 FOREIGN Foreign worker Binary 0: No, 1: Yes

32 RESPONSE Credit rating is Binary 0: No, 1: Yes

good

The consequences of misclassification have been assessed as follows: the costs of a false
positive (incorrectly saying an applicant is a good credit risk) outweigh the cost of a false
negative (incorrectly saying an applicant is a bad credit risk) by a factor of five. This can be
summarized in the following table.

111
 Introduction to Statistics and Probability

Predicted (Decision)

Actual Good Bad (Reject)

(Accepted)

Good 0 100DM

Bad 500DM 0

Table 1.3 Opportunity Cost Table (in deutch Marks)

The opportunity cost table was derived from the average net profit per loan as shown below:

Predicted (Decision)

Actual Good (Accepted) Bad (Reject)

Good 100DM 0

Bad -500DM 0

Table 1.4 Average Net Profit

Let us use this table in assessing the performance of the various models because it is simpler
to explain to decision-makers who are used to thinking of their decision in terms of net profits.

1. Review the predictor variables and guess from their definition at what their role might
be in a credit decision. Are there any surprises in the data?

2. Divide the data randomly into training (60%) and validation (40%) partitions, and
develop classification models using the following data mining techniques in SPSS:

• Classification trees
• Neural networks
• Discriminant Analysis.

Grading Scheme
The assignment is marked out of 50.

Answer
Solutions and answers will be provided by the instructor.

112
 Unit 4: Applications of Probability and Statistics in ACS

Unit Readings and Other Resources

The readings in this unit are to be found at course level readings and other resources.

Module Summary
In modern computer science, software engineering, and other fields, the need arises to
make decisions under uncertainty. Presenting probability and statistical methods, simulation
techniques, and modeling tools, Probability and Statistics for applied Computer Science helps
students solve problems and make optimal decisions in uncertain conditions, select stochastic
models, compute probabilities and forecasts, and evaluate performance of computer systems
and networks.

After introducing probability and distributions, this easy-to-follow module provides two
course options. The first approach is a probability-oriented course that begins with stochastic
processes, Markov chains, and queuing theory, followed by computer simulations and Monte
Carlo methods. The second approach is a more standard, statistics-emphasized course that
focuses on statistical inference, estimation, hypothesis testing, and regression. The Module is
illustrated throughout with numerous examples, exercises, figures, and tables that stress direct
applications in computer science and software engineering.

By the end of this course, advanced undergraduate and beginning graduate students should
be able to read a word problem or a corporate report, realize the uncertainty involved in the
described situation, select a suitable probability model, estimate and test its parameters based
on real data, compute probabilities of interesting events and other vital characteristics, and
make appropriate conclusions and forecasts.

113
 Introduction to Statistics and Probability

Module Course Assessment

Identify the choice that best completes the statement or answers the question.

1. A random sample of 1000 people was taken. Four hundred fifty of the people in the
sample favored Candidate

A. The 95% confidence interval for the true proportion of people who favors Candidate A is

a. 0.419 to 0.481

b. 0.40 to 0.50

c. 0.45 to 0.55

d. 1.645 to 1.96

In order to estimate the average time spent on the computer terminals per student at a
local university, data were collected for a sample of 81 business students over a one-week
period. Assume the population standard deviation is 1.8 hours.

____ 2. Refer to Exhibit 8-1. With a 0.95 probability, the margin of error is approximately

a. 0.39

b. 1.96

c. 0.20

d. 1.64

6. The probability of committing a Type I error when the null hypothesis is true is

a. the confidence level

b. b

c. greater than 1

d. the Level of Significance

7. The probability of making a Type I error is denoted by

a. a

b. b

c. 1 - a

d. 1 - b

114
 Unit 4: Applications of Probability and Statistics in ACS

8. For a one-tailed test (upper tail), a sample size of 26 at 90% confidence, t =

a. 1.316

b. -1.316

c. -1.740

d. 1.740

9. For a one-tailed test (lower tail) with 22 degrees of freedom at 95% confidence, the
value of t =

a. -1.383

b. 1.383

c. -1.717

d. -1.721

Problem

A random sample of 49 lunch customers was taken at a restaurant. The average amount of
time the customers

in the sample stayed in the restaurant was 45 minutes with a standard deviation of 14
minutes.

a. Compute the standard error of the mean.

b. With a .99 probability, what statement can be made about the size of the margin of
error?

c. Construct a 99% confidence interval for the true average amount of time customers
spent in

The monthly incomes from a random sample of workers in a factory are shown below.

115
 Introduction to Statistics and Probability

Monthly Income

(In $1,000)

4.0

5.0

7.0

4.0

6.0

6.0

7.0

9.0

a. Compute the standard error of the mean (in dollars).

b. Compute the margin of error (in dollars) at 95% confidence.

c. Compute a 95% confidence interval for the mean of the population. Assume the
population has a normal distribution. Give your answer in dollars.

The proprietor of a boutique in New York wanted to determine the average age of his
customers. A random sample of 53 customers revealed an average age of 28 years with
a standard deviation of 4 years. Determine a 98% confidence interval estimate for the
average age of all his customers.

A coal company wants to determine a 95% confidence interval estimate for the average
daily tonnage of coal that they mine. Assuming that the company reports that the standard
deviation of daily output is 200 tons, how many days should they sample so that the margin
of error will be 39.2 tons or less?

The average score of a sample of 87 senior business majors at UTC who took the Graduate
Management Admission Test was 510 with a standard deviation of 36. Provide a 98%
confidence interval for the mean of the population.

116
 Unit 4: Applications of Probability and Statistics in ACS

In order to determine the average weight of carry-on luggage by passengers in airplanes, a

sample of 25 pieces of carry-on luggage was collected and weighed. The average weight
was 18 pounds. Assume that we know the standard deviation of the population to be 7.5
pounds.

a. Determine a 97% confidence interval estimate for the mean weight of the carry-on
luggage.

b. Determine a 95% confidence interval estimate for the mean weight of the carry-on
luggage.

In order to determine the average price of hotel rooms in Atlanta, a sample of 64 hotels
was selected. It was determined that the average price of the rooms in the sample was
$108.50 with a standard deviation of $16. a. Formulate the hypotheses to determine
whether or not the average room price is significantly different from $112.

b. Compute the test statistic.

c. At 95% confidence using the p-value approach, test the hypotheses. Let a = 0.1.

Course References
Optional readings and other resources:

• http://mathworld.wolfram.com/Probability: Wolfram is a useful site that provides

insights in number theory while providing new challenges and methodology in
number theory.
• http://en.wikipedia.org/wiki/Probability: Mathsguru is a website that helps
learners to understand various branches of number theory module.

Required readings and other resources:

• DeGroot, Morris H., and Mark J. Schervish. Probability and Statistics. 3rd ed.
Boston, MA: Addison-Wesley, 2002. ISBN: 0201524880.
• John A. Rice, Mathematical Statistics and Data Analysis (with CD Data Sets)
(Duxbury Advanced). 3rd Edition, Cengage Learning, 2006, ISBN-13 978-
0534399429 for reading.
• http://ocw.mit.edu/courses/mathematics/18-05-introduction-to-probability-and-
statistics-spring-2005/index.htm for self practicing. Lots of interesting exercises.
• http://books.google.com/books/about/
Probability_and_Random_Processes_With_Ap.htm
• http://en.wikipedia.org/wiki/Probability: Mathsguru is a website that helps
learners to understand various branches of number theory module. It is easy to
access through Google search and provides very detailed information on various
probability questions.

117
 Introduction to Statistics and Probability

• http://en.wikipedia.org/wiki/Naive_Bayes_classifier
• http://ocw.mit.edu/courses/sloan-school-of-
management/15-062-data-mining-spring-2003/
lecture-notes/
• http://archive.ics.uci.edu/ml/

118
The African Virtual University
Headquarters

Cape Office Park

Ring Road Kilimani

PO Box 25405-00603

Nairobi, Kenya

Tel: +254 20 25283333

[email protected]

[email protected]

The African Virtual University Regional

Office in Dakar

Université Virtuelle Africaine

Bureau Régional de l’Afrique de l’Ouest

Sicap Liberté VI Extension

Villa No.8 VDN

B.P. 50609 Dakar, Sénégal

Tel: +221 338670324

[email protected]