Machine learning

Machine Learning
Tobi Ogunbiyi & Alex Vermeulen
1

What is Machine Learning?
A computer program is said to learn if its
measured performance on a task improves with
experience.
2

• Google’s self-driving car
• Optical Character Recognition (OCR)
• Google street view
• Facebook
Machine Learning Applications
3

Supervised Learning
The machine is “trained” using examples for
which we know the correct answer.
- Labeled data
- Used for classiﬁcation or prediction
4

• Features: shape, size, colour, and sound
• Labels: “cow”, “pig”, “chicken”, “llama”
Supervised Learning
Example
5

Unsupervised Learning
Tries to find patterns and groupings by analyzing
the characteristics of the data
• Unlabeled data
• Identiﬁes patterns and groupings in the data
6

• No labels, just looking to group similar animals
• Features: shape, size, colour, and sound
Unsupervised Learning
7

Cloud accounting software designed for small,
service-based businesses
8

9
330
expenses per month
8
hours per year
Story Telling
7
seconds per expense
-

How can Machine Learning
Help?
• Simple & Painless
• Create a better user experience
• Save customers time and let them do what they
do best!
10

12
Data Collection
Pre-Processing
Sampling
Build Model
Refinement
Evaluation
Prediction

12
Data Collection
Pre-Processing
Sampling
Build Model
Refinement
Evaluation
Prediction
Categorized expenses from the last year

13
#5795# QTH Toronto ON
#991# Toronto ON
Roots #130 Etobicoke ON
True North Climbing Toronto ON
Tim Hortons
Eddie Bauer Canada
M9C
Pre Processing

13
QTH Toronto ON
Toronto ON
Roots Etobicoke ON
True North Climbing Toronto ON
Tim Hortons
Eddie Bauer Canada
Pre Processing

14
Tim Hortons QTH Toronto ON
Vectorize
(i)
(ii) Eddie Bauer Canada Toronto ON

14
Vectorize
“tim”
“hortons”
“qth”
“toronto”
“on”
“eddie”
“bauer”
“canada”
(i)

14
Vectorize
“tim”
“hortons”
“qth”
“toronto”
“on”
“eddie”
“bauer”
“canada”
(i)
(i)
-
-
-
-
-
-
-
-

14
Vectorize
“tim”
“hortons”
“qth”
“toronto”
“on”
“eddie”
“bauer”
“canada”
(i)
(i) 1
-
-
-
-
-
-
-

14
Vectorize
“tim”
“hortons”
“qth”
“toronto”
“on”
“eddie”
“bauer”
“canada”
(i)
(i)
1
1
-
-
-
-
-
-

14
Vectorize
“tim”
“hortons”
“qth”
“toronto”
“on”
“eddie”
“bauer”
“canada”
(i)
1
1
1
1
1
0
0
0
(i)

14
Vectorize
“tim”
“hortons”
“qth”
“toronto”
“on”
“eddie”
“bauer”
“canada”
(i)
1
1
1
1
1
0
0
0
(i) (ii)
-
-
-
-
-
-
-
-

14
Vectorize
“tim”
“hortons”
“qth”
“toronto”
“on”
“eddie”
“bauer”
“canada”
(i)
1
1
1
1
1
0
0
0
(i) (ii)
-
-
-
-
-
1
-
-

14
Vectorize
“tim”
“hortons”
“qth”
“toronto”
“on”
“eddie”
“bauer”
“canada”
(i)
1
1
1
1
1
0
0
0
(i) (ii)
0
0
0
1
1
1
1
1

15
Transform
Term Frequency - Inverse Document Frequency
= term frequency x inverse document freq.

15
Transform
A numerical statistic that reflects how important
or descriptive a term is to a single document in a
collection.

15
Transform
term freq. = occurrences of term in document

15
Transform
inverse document freq.= log
total documents
docs containing the term
term freq. = occurrences of term in document

16
Tf-Idf Example
“tim”
“hortons”
“qth”
“toronto”
“on”
“eddie”
“bauer”
“canada”
(i)
1
1
1
1
1
0
0
0
(i)

16
Tf-Idf Example
tf(tim, d1) = occurrences of term
“tim”
“hortons”
“qth”
“toronto”
“on”
“eddie”
“bauer”
“canada”
(i)
1
1
1
1
1
0
0
0
(i)

16
Tf-Idf Example
“tim”
“hortons”
“qth”
“toronto”
“on”
“eddie”
“bauer”
“canada”
(i)
1
1
1
1
1
0
0
0
(i) tf(tim, d1) = 1

16
Tf-Idf Example
“tim”
“hortons”
“qth”
“toronto”
“on”
“eddie”
“bauer”
“canada”
(i)
1
1
1
1
1
0
0
0
(i)
idf(tim, D) = log
total docs
docs cont. term
tf(tim, d1) = 1

16
Tf-Idf Example
“tim”
“hortons”
“qth”
“toronto”
“on”
“eddie”
“bauer”
“canada”
(i)
1
1
1
1
1
0
0
0
(i)
idf(tim, D) = log
docs cont. term
2
tf(tim, d1) = 1

16
Tf-Idf Example
“tim”
“hortons”
“qth”
“toronto”
“on”
“eddie”
“bauer”
“canada”
(i)
1
1
1
1
1
0
0
0
(i)
idf(tim, D) = log
2
1
tf(tim, d1) = 1

16
Tf-Idf Example
= 0.301
“tim”
“hortons”
“qth”
“toronto”
“on”
“eddie”
“bauer”
“canada”
(i)
1
1
1
1
1
0
0
0
(i)
idf(tim, D) = log
2
1
tf(tim, d1) = 1

16
Tf-Idf Example
tﬁdf(tim, d1) = 1 x 0.301 = 0.301
= 0.301
“tim”
“hortons”
“qth”
“toronto”
“on”
“eddie”
“bauer”
“canada”
(i)
1
1
1
1
1
0
0
0
(i)
idf(tim, D) = log
2
1
tf(tim, d1) = 1

16
Tf-Idf Example
“tim”
“hortons”
“qth”
“toronto”
“on”
“eddie”
“bauer”
“canada”
(i)
1
1
1
1
1
0
0
0
(i) .301
.301
.301
0
0
0
0
0

17
Data Collection
Pre-Processing
Sampling
Build Model
Refinement
Evaluation
Prediction

17
Data Collection
Pre-Processing
Sampling
Build Model
Refinement
Evaluation
Prediction
80% training data 
20% testing data

• Multinomial Logistic Regression (supervised)
• Used for unordered, categorical outputs with
more than 2 possible categories
• Series of linear sub-models which output a real
number
18
Classify

Eggwidth
2
2.5
3
3.5
4
4.5
Egg length
4 4.5 5 5.5 6 6.5 7
Duck egg
Chicken egg
Linear regression

Linear regression
Eggwidth
2
2.5
3
3.5
4
4.5
Egg length
4 4.5 5 5.5 6 6.5 7
Duck egg
Chicken egg

20
Logistic Function
• Normalizes the output of the model to a score
between 0 and 1

20
Logistic Function
1 + e-x
logistic function =
1
between 0 and 1

Output
0
0.25
0.5
0.75
1
Input
-5 -4 -3 -2 -1 0 1 2 3 4 5
20
Logistic Function
between 0 and 1

21
Data Collection
Pre-Processing
Sampling
Build Model
Refinement
Evaluation
Prediction

Validation
• Tested against the 20% reserved testing data set
• Cross validation against a completely separate
set of expense data
22

23
Data Collection
Pre-Processing
Sampling
Build Model
Refinement
Evaluation
Prediction

Reﬁnement
• How do you know if your model is good enough?
• What can you do to improve your model?
• Adjust the amount of data
• Clean irrelevant parts of the data
• Tweak parameters of the algorithm
24

MeanAccuracy(%)
0%
10%
20%
30%
40%
50%
60%
70%
80%
90%
100%
Sample size
10K 100K 200K 300K 400K 500K
SVM
Multinomial Naive Bayes
Bernoulli Naive Bayes
Logistic Regression L1
Experiment

26
Data Collection
Pre-Processing
Sampling
Build Model
Refinement
Evaluation
Prediction

• Use the model to classify new, unlabeled data
Prediction
27

Prediction
27
Tim Hortons 3335 QPS Toronto ON

Prediction
27
“Advertising”
“Car & Truck Expenses
“Meals & Entertainment”
“Personal”
“Rent or Lease”
“Travel”
“Utilities”

Prediction
27
“Advertising”
“Personal”
“Rent or Lease”
“Travel”
“Utilities”
0.020
0.018
0.710
0.230
0.003
0.011
0.008

Prediction
27
“Advertising”
“Personal”
“Rent or Lease”
“Travel”
“Utilities”
0.020
0.018
0.710
0.230
0.003
0.011
0.008
Threshold: 0.6

Getting Started
• What kind of data do you have?
• Labeled or unlabeled?
• What questions are you trying to answer?
• Make predictions
• Label or classify
• Identify patterns or groupings
28

Lessons Learned
• Machines are intelligent, but not magicians
• It’s easy to know you’re wrong, but harder to
know when you’re right
• Some people prefer to have control
29

Take away?
• Opens up new opportunities
• Potential to deliver amazing user experiences
• Machine Learning is fun!
30

Resources
• Interested in following along with FreshBooks
Learnings?
• medium.com/@freshbookspd
• Want to learn more about Machine Learning?
• udacity.com
• coursera.org
• Python’s scikit-learn
32

More examples of our
experimentation…
33

Model Build TimeTimetotrain(min)
0
1
2
3
4
5
6
7
8
9
10
Sample size
10K 100K 200K 300K 400K 500K
SVM

Prediction TimeTimetopredict(ms)
0ms
10ms
20ms
30ms
40ms
50ms
60ms
70ms
80ms
Sample size
10K 100K 200K 300K 400K 500K
SVM

File Size
FileSize(MB)
0MB
500MB
1,000MB
1,500MB
2,000MB
2,500MB
Sample size
10K 100K 200K 300K 400K 500K
SVM

Machine learning

More Related Content

Viewers also liked (13)

Similar to Machine learning (20)

Recently uploaded (20)

Machine learning