EE555 Machine Learning (3 credit hours)

cobblerbeggarΤεχνίτη Νοημοσύνη και Ρομποτική

15 Οκτ 2013 (πριν από 3 χρόνια και 6 μήνες)

83 εμφανίσεις

Course Outline


EE
555

Machine Learning


(3 credit hours)


Fall

200
5


Schedule

Lecture
s

on
Tue/Thu 5:00


6:15
pm

Website

http://sst.imt.edu.pk

Instructor

Haroon Atique Babri

Contact

babri@umt.edu.pk

Ext. 323

Office

1st Floor, SST, UMT

52
-
L Gulberg III

Office
Hours

Appointment

Teaching
Assistant

--

Contact

--

Office

--

Office
Hours

--


Course
Description

This course is a broad introduction to machine learning and includes discussions of
most of the major approaches currently being investigated. We wi
ll discuss general
issues in machine learning, as well as presently well known algorithms. We will also
compare and contrast the various approaches, determining under which conditions
each is most appropriate.


Expected
Outcomes

Upon completion of this
course, students will:



Gain insight into the field of
machine learning and the type of problems it is best
suited to



Understand
the need for different approaches and their limitations and how they
compare with each other



Gain experience in implementing mac
hine learning algorithms
.

Textbook

References



Introduction to machine learning by Ethem Alpaydin, MIT Press, 2004



Machine Learning by
Tom
Mitchell, McGraw Hill, 1997



Pattern Classification by Duda, Hart and Stork, 2003

Grading
Policy



Quizzes 20%



Assignme
nts
:
15
%



Midterm
:
30
%



Final: 35%


Lecture Plan



Lec
ture

Topic


Reading


----------------------------------------------------------------------------------------------------------------


1
-
2


Introduction


Chap1



Overview of the field


3
-
5


Supervised
learning


Chap2



VC dimension, PAC learning



Regression



Model selection and Generalization


6
-
8


Bayesian Decision Theory

Chap3



Losses and Risks



Discriminant functions



Bayesian Networks


9



12

Parametric methods


Chap 4
, 5



Maximum likelihood
estimation



Parameter estimation



Multivariate classification



Multivariate regression



1
3


15

Dimensionality reduction

Chap 6



Principal Components Analysis



Factor Analysis



Multidimensional Scaling



Linear Discriminant Analysis


16


MIDTERM


1
7


18


Clustering

Chap 7



k
-
means, EM



19


21


Nonparametric methods

Chap 8



Nearest neighbor estimators



Classification

and Regression


21


22

Decision Trees


Chap 9



Pruning



Rule extraction


23


26

Li
near Discrimina
tion


Chap 10



Gradient descent



Logistic discrimination



Support vector machines


27


30

Comparing Classification Algorithms

Chap 14



Cross
-
validation and resampling



Error measurement



Hypothesis testing



Evaluating algorithm perfo
rmance