Artificial Intelligence Application

By Dr. Julius Aroba

 Understand the basics of Machine learning and why/how
machine learning methods are used to learn from data to
find underlying patterns for prediction and decision-making.

Lesson The applications of Machine Learning and research fields.

Goals Understand the basics and concept of the learning problem.

Differentiate between supervised, unsupervised learning and

Reinforcement learning.
 R. Duda, P. Hart & D. Stork, Pattern Classification (2nd ed.),
Wiley

Source T. Mitchell, Machine Learning, McGraw-Hill

Materials Christopher Bishop, Neural Networks for Pattern Recognition,

Oxford University Press, 1995

Introduction to Machine Learning by Ethem Alpaydin (2nd

edition, 2010, MIT Press).

Foundations of Machine Learning by Afshin Rostamizadeh,

Ameet Talwalkar, and Mehryar Mohri (2012, MIT Press)
 What AI can do for us

Endless possibilities
in daily applications, Recognize objects in Translate between Speak, transcribe
unbiased decisions, images languages speech
digital assistance, etc.

The list is very large

Recognize/detect
Recognize emotions Drive, fly a drone, … in all areas of
diseases, diagnostics
knowledge
 Machine Learning Basics
• Machine Learning is a branch of Artificial Intelligence, which focuses
on methods that learn from data.
• These uncovered patterns are then used to predict future data, or to
perform other kinds of decision-making under uncertainty.
• The key premise is learning from data!!

Machine
Labeled Data Learning
algorithm
Training
Prediction

Learned
Labeled Data Prediction
model
Machine Learning Basics (cont.)

• Addresses the problem of analyzing huge bodies of data so that

they can be understood.
• Providing techniques to automate the analysis and exploration
of large, complex data sets.
• Tools, methodologies, and theories for revealing patterns in
data – critical step in knowledge discovery.
Popular quotes on Machine Learning

“Breakthroughs in machine Machine learning is the hot

“Machine learning is the next
learning would be worth ten new thing”
Internet”
Microsofts” (Bill Gates, (John Hennessy, President,
(Tony Tether, Director, DARPA)
Chairman, Microsoft) Stanford)

“Web rankings today are mostly “Machine learning is going to

“Machine learning is today’s
a matter of machine learning” result in a real revolution” (Greg
discontinuity”
(Prabhakar Raghavan, Dir. Papadopoulos, Former CTO,
(Jerry Yang, Founder, Yahoo)
Research, Yahoo) Sun)

“Machine learning today is one

of the hottest aspects of
computer science” (Steve
Ballmer, CEO, Microsoft)
Big Data is
Everywhere

We are in the era of big data!

40 billion indexed web pages
100 hours of video are uploaded to
YouTube every minute
The deluge of data calls for
automated methods of data
analysis, which is what machine
learning provides
What Machine • Automating automation
Learning brings to the • Getting computers to program themselves.
table • Let the data do the work instead!
Examples of ML
applications

Machine learning plays a key role in many areas of science,

finance and industry:
 Predict whether a patient, hospitalized due to a heart attack,
will have a second heart attack.
 The prediction is to be based on demographic, diet and
clinical measurements for that patient.
 Predict the price of a stock in 6 months from now, on the
basis of company performance measures & economic data.
 Identify the numbers in a handwritten ZIP code, from a
digitized image.
 Estimate the amount of glucose in the blood of a diabetic
person, from the infrared absorption spectrum of that
person’s blood.
 Identify the risk factors for prostate cancer, based on clinical
and demographic variables.
 Statistics / Statistical Learning
ML
Research Data Mining
Fields
Pattern Recognition

Business Intelligence

Signal Processing
 Data
• NIPS 2003 feature selection contest
• mldata.org
• UCI machine learning repository
Machine
Contests
learning
• Kaggle
resources
Software
• Python sci-kit
• R
• Tensorflow
• Your own code
The Learning
Problem

• Learning from data is used in

situations where we don’t have any
analytic solutions, but we do have
data that we can use to construct
an empirical solution.
• The basic premise of learning from
data is the use of a set of
observations to uncover an
underlying process.
The Learning Problem (cont.)
The
Learning
Problem
(cont.)
Types of Learning

• Supervised (inductive) learning

– Training data includes desired outputs
• All the predictors, Xi, and the response, Yi, are observed.
• Regression and classification methods
• Unsupervised learning
– Training data does not include desired outputs
• Here, only the Xi’s are observed (not Yi’s).
• We need to use the Xi’s to guess what Y would have been, and then build a model form
there.
• Clustering and principal components analysis
• Reinforcement learning
– Rewards from sequence of actions
Supervis
ed
Learning
Supervised
Learning

• Supervised learning categories and techniques

• Numerical classifier functions

• Linear classifier, perceptron, logistic regression, support
vector machines (SVM), neural networks
• Parametric (probabilistic) functions
• Naïve Bayes, Gaussian discriminant analysis (GDA), hidden
Markov models (HMM), probabilistic graphical models

• Non-parametric (instance-based) functions

• k-nearest neighbors, kernel regression, kernel density
estimation, local regression
• Symbolic functions
• Decision trees, classification and regression trees (CART)

• Aggregation (ensemble) learning

• Bagging, boosting (Adaboost), random forest
 Supervised
Learning:
Examples
Handwritten Digit
Recognition:
Identify single
digits 0~9 based
on images
Supervised Learning:
Examples

Face
Detection/Recognition:
Identify human faces
Supervised Learning: Training
Supervised
Learning:
Testing
Supervised
Learning:
Testing
Unsupervised learning
• Unsupervised learning categories and techniques
• Clustering
• k-means clustering
• Mean-shift clustering
• Spectral clustering
• Density estimation
• Gaussian mixture model (GMM)
• Graphical models
• Dimensionality reduction
• Principal component analysis (PCA)
• Factor analysis
Unsupervised learning
• Unsupervised learning categories and techniques
• Clustering
• k-means clustering
• Mean-shift clustering
• Spectral clustering
• Density estimation
• Gaussian mixture model (GMM)
• Graphical models
• Dimensionality reduction
• Principal component analysis (PCA)
• Factor analysis
Unsupervised Learning
Unsupervised Learning (cont.)

• The training data does not contain any output information at all (i.e.
unlabeled data).

• Viewed as the task of spontaneously finding patterns and structure in

input data.

• Viewed as a way to create a higher-level representation of the data

and dimension reduction.
Reinforcement Learning

• Main idea: Learning with a Delayed Reward

• Uses dynamic programming and supervised learning

• Addresses problems that can not be addressed by

• regular supervised methods
• E.g., Useful for Control Problems.

• Dynamic programming searches for optimal policies

Summary

• Overview of Artificial Intelligence and what it can do for us

• Basics of Machine learning
• Applications of Machine Learning
• Key concepts of the learning problem
• Supervised, unsupervised and reinforcement learning