18-794-Carnegie Mellon University Africa - Carnegie Mellon University

18-794

18-794

Pattern Recognition Theory

Course discipline: Data Science

Elective
Units: 12
Lecture/Lab/Rep hours/week: 3 Lecture hours per week

Semester/year offered (fall/spring, even/odd/all years): Spring
Pre-requisites: Some programming experience is needed

Will help if students have taken following courses:

Data structures, algorithms (15-211 or similar a must, 15-451 or similar useful but not required).

Digital Signal Processing, Linear algebra, Probability and Statistics (36-217 or similar is useful but not required)

Course description:

The course will be relevant for students interested in machine learning as applied to computer vision and image processing. This will be combined with practical applications, exposure to industry-standard tools, and research techniques.

Learning objectives:

Successful application of in-class theory to solving home-work assignments, mid-term exams, and surprise in-class quizzes. Ability to understand a variety of research topics, and present these to an audience. Ability to interpret the requirement of a complex, real-life problem and solve it using machine learning.  

Outcomes:

After completing this course, students should be able to:

Understand the basics of machine learning algorithms from first principles. Ability to apply techniques to solving real-world problems. Understand a research paper completely, present complex problems in front of an audience. 

Content Details:

Lecture (90 mins)

Week number

Date

Description

Note

Homeworks

1

1

11-Jan-15

Representation of images, noise and sampling.

2

1

11-Jan-15

Linear algebra Intro, PCA

HW-1 Handed out

3

2

18-Jan-15

SVD, Gradient descent, function optimization. Discuss modalities for semester projects.

No classes in PIT on Jan 18

4

2

18-Jan-15

--Topic Continues--

5

3

25-Jan-15

Parametric models (Logistic, Gaussian, GMM)

HW-1 Due

6

3

25-Jan-15

--Topic Continues--

7

4

1-Feb-15

Morphological operators,  FFT, Wavelets, Feature Extraction

No classes in KGL on Feb 1

HW-2 Handed out

8

4

1-Feb-15

--Topic Continues--

9

5

8-Feb-15

Neural networks 1

10

5

8-Feb-15

Neural networks 2

Project title, research reading groups due

11

6

15-Feb-15

Neural networks 3

12

6

15-Feb-15

Neural networks 4

HW-2 Due

13

7

22-Feb-15

K-Nearest neighbor

Feb 22-29 Mini-3 Faculty Course Evaluations

14

7

22-Feb-15

Clustering

15

8

1-Mar-15

No Classes

Feb 29-Mar 3 No classes Graduate Mini-3 Exam Days

16

8

1-Mar-15

No Classes

 

Research topic selected papers due

17

9

8-Mar-15

No Classes

Mar 7-11, Spring Break; No Classes in Pittsburgh

18

9

8-Mar-15

No classes

 

19

10

15-Mar-15

Support Vector Machines

HW-3 handed out

20

10

15-Mar-15

--Topic Continues--

21

11

22-Mar-15

Hidden Markov Models

March 25 Good Friday ( No classes in KGL)

22

11

22-Mar-15

--Topic Continues--

23

12

29-Mar-15

Decision Trees, Random forests

24

12

29-Mar-15

Boosting

HW-3 Due

25

13

5-Apr-15

No Classes

Apr 7-13 Genocide memorial week ( No classes in KGL)

26

13

5-Apr-15

No Classes

27

14

12-Apr-15

Research topic presentations

Research topic presentations on selected papers

28

14

12-Apr-15

Research topic presentations

Research topic presentations on selected papers

29

15

19-Apr-15

Genetic Algorithms

30

15

19-Apr-15

--Topic Continues--

31

16

26-Apr-15

RANSAC, Stereo Correspondence

32

16

26-Apr-15

--Topic Continues--

33

17

3-May-15

May 2 Labor day ( No classes in KGL)

34

17

3-May-15

 

18

10-May-15

Projects due

May 7-9, 10th onwards Reading days and final exam week

 

18

10-May-15