Scribd
Upload
32 views7 pages

Business and Economic Statistics: 10. Predictive Analytics - Estimation

This document discusses predictive analytics and statistical inference. It notes that descriptive analytics can discover patterns in data, like differences in average doctor visits between males (1.871) and females (2.682). However, some differences are small and could be due to chance rather than real effects. Predictive analytics uses statistical inference to determine whether differences are real or due to random variation. Statistical inference uses sample statistics to estimate population parameters, but there is uncertainty since samples vary. This uncertainty is measured using confidence intervals, hypothesis testing, and standard errors.

Uploaded by

Mohammad Rashman
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PPTX, PDF, TXT or read online on Scribd
32 views7 pages

Business and Economic Statistics: 10. Predictive Analytics - Estimation

This document discusses predictive analytics and statistical inference. It notes that descriptive analytics can discover patterns in data, like differences in average doctor visits between males (1.871) and females (2.682). However, some differences are small and could be due to chance rather than real effects. Predictive analytics uses statistical inference to determine whether differences are real or due to random variation. Statistical inference uses sample statistics to estimate population parameters, but there is uncertainty since samples vary. This uncertainty is measured using confidence intervals, hypothesis testing, and standard errors.

Uploaded by

Mohammad Rashman
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PPTX, PDF, TXT or read online on Scribd
You are on page 1/ 7

ETC1000

BUSINESS AND ECONOMIC


STATISTICS
10. Predictive Analytics
Estimation

PREDICTIVE ANALYTICS:
DATA AND DIFFERENCES

Consider data on number of doctors visits people


make.

PREDICTIVE ANALYTICS:
REAL DIFFERENCES VS CHANCE DIFFERENCES

Descriptive

Analytics allows us to discover a number of


patterns and relationships:

male

= 1.871, female = 2.682

Low-income = below $1,800pm, high above $1,800pm.


low = 2.272, high = 2.289
Proportion with no doctors visits:
Full sample: 38.0%
Winter: 37.9%
Other Seasons: 38.1%

Correlation (Doctors visits & Education) = -.092


Correlation (Doctors visits & Income) = -.005

PREDICTIVE ANALYTICS:
REAL DIFFERENCES VS CHANCE DIFFERENCES
Some differences in means and proportions are small,
and some correlations are small; others do not seem
as small.
Small differences / values are probably just zero, and
the non-zero value was due to random variation. But
how do we decide what is small?
Need to identify real / substantive differences vs
small, coincidental differences.
Real, substantive differences are likely to persist, so
relevant to decisions & predicting behaviour.

Predictive Analytics requires ideas of Statistical


Inference.

PREDICTIVE ANALYTICS:
STATISTICAL INFERENCE

Statistical Inference : use a sample of data to


estimate characteristics of a population.

PREDICTIVE ANALYTICS:
STATISTICAL INFERENCE
A Sample Statistic is used
as an estimator of a Population Parameter
Population Parameter

Sample Statistic

Mean

Sample Mean

Proportion

Sample Proportion p

Difference between means

Difference between sample means

Difference in proportions

Difference in sample proportions

Correlation

Sample Correlation r

etc

etc

PREDICTIVE ANALYTICS:
STATISTICAL INFERENCE

Whenever we use a sample statistic to estimate a


population parameter, there is uncertainty: the
estimate is only an approximation to the true
value.
If we take a new sample, we will get a different
estimate. Both cannot be correct!
The uncertainty is captured with ideas of
confidence intervals, hypothesis testing, standard
errors.

You might also like