You are on page 1of 17

Machine Learning with MATLAB

Leuven Statistics Day2014


Rachid Adarghal, Account Manager
Jean-Philippe Villaréal, Application Engineer

© 2014 The MathWorks, Inc.1


Side note: Design of Experiments with
MATLAB

2
What You Will Learn

 Get an overview of Machine Learning

 Machine learning models and techniques available


in MATLAB

 MATLAB as an interactive environment


– Evaluate and choose the best algorithm

3
Machine Learning

 Characteristics
– Lots of data (many variables)
– System too complex to understand
the governing equation

4
Domains of Application

 Handwriting recognition
 Autonomous vehicles
 DNA sequencing / Genomics
 Cancer tumor classification
 Social Network Analysis
 Astronomical Data Analysis
 Market Segmentation
 Organizing Computer Cluster for efficiency
 Spam / non spam email classification
 Hearing headsets: optimizing signal (Cocktail party)
 Shazam / SoundHound
 FingerPrinting

5
Challenges – Machine Learning

 Lots of data, with many variables (predictors)

 Data is too complex to know the governing


equation

 Significant technical expertise required


– Black box modelling

 No “one size fits all” approach: Requires an


iterative approach:
– Try multiple algorithms, see what works best
– Time consuming to conduct the analysis
– Know-how required to debug your algorithm efficiently
6
MATLAB Solutions

 Strong environment for interactive exploration

 Algorithms and Apps to get started


– Clustering, Classification, Regression
– Neural Network app, Curve fitting app

 Easy to evaluate, iterate, and choose the best


algorithm

 Parallel Computing

 Deployment for Data Analytics workflows

7
Overview – Machine Learning

Type of Learning Categories of Algorithms

Unsupervised
Clustering
Learning

Group and interpret


data based only Recommender
Machine
on input data systems
Learning

Classification
Supervised
Learning

Develop predictive Regression


model based on both
input and output data

9
Unsupervised Learning
k-Means

Self-Organizing
Maps
Partitional
Clustering
Hierarchical
clustering
Clustering
Overlapping Fuzzy C-Means

Clustering
Gaussian
Mixture

Hidden Markov
Model

10
Supervised Learning

Regression

Neural Ensemble Non-linear Reg. Linear


Decision Trees
Networks Methods (GLM, Logistic) Regression

Classification

Support Vector Discriminant Nearest


Naive Bayes
Machines Analysis Neighbor

11
Supervised Learning - Workflow

Speed up Computations

Select Model

Data Train the Model Use for Prediction


Import Data
Known data Model
Model Predicted
Explore Data
Responses
Prepare Data Known responses New Data

Measure Accuracy
12
Example – Bank Marketing Campaign

 Goal:
– Predict if customer would subscribe to
bank term deposit based on different 100
Bank Marketing Campaign
Misclassification Rate

attributes
90

80

70

60

Percentage
No
Misclassified
50
Yes

 Approach: 40

30
Misclassified

– Train a classifier using different models


20

10


0

Measure accuracy and compare models

B
or s

s
et

s
VM
n

ge
is

r ee
aye

dT
s sio
lN

ly s

ag
hb

nT
o rt

ce
ura

B
na

eB
g re

ig

ive

pp

du
Ne
A

io
Ne

Tre
Re

cis
Su

Re
nt

Na
t
res
ina

De
tic
gis

cr im

ea
k -n

Lo

Dis
Reduce model complexity
– Use classifier for prediction

Data set downloaded from UCI Machine Learning repository


http://archive.ics.uci.edu/ml/datasets/Bank+Marketing
13
Summary– Bank Marketing Campaign

 Numerous predictive models with 100

90
Bank Marketing Campaign
Misclassification Rate

rich documentation 80

70

– Clustering, regression, classification


60

Percentage
No
Misclassified
50
Yes
Misclassified
40

30

20

10

 Interactive tools to help discovery 0

B
or s

s
et

s
VM
n

ge
is

r ee
aye

dT
s sio
lN

ly s

ag
hb

nT
o rt

ce
ura

B
na

eB
g re

ig

ive

pp

du
Ne
A

io
Ne

Tre
Re

cis
Su

Re
nt

Na
– Histograms, bar charts, ROC curves

t
res
ina

De
tic
gis

cr im

ea
k -n
Lo

Dis
– Graphical Apps

 Built-in parallel computing support

 Quick prototyping
– Focus on modeling not programming
14
Learn More: Machine Learning with MATLAB

Visit our discovery page: www.mathworks.com/machine-learning

15
Deploying / Sharing Your Application
APPS

Builder
Web
NE
MATLAB Compiler Builder
Web
JA
.dll
.lib .exe
Builder
.CTF Ex

MATLAB Coder
Web

MEX

.exe .lib .dll

16
MathWorks Services

 Trainings:
– More that 30 course offerings
 Consulting Services
– Enhance your team
 Technical Support
– Ask questions
 An active community:
– MATLAB Central
– File exchange
– Blogs
– Newsletters

17
Thank you for attending!

© 2014 The MathWorks, Inc.


18

You might also like