###EE 636 – MATRIX COMPUTATIONS

Semester: Spring 2016

Instructor: Harish K. Pillai

Motivation: In majority of the computations while solving engineering problems, we come across with the situation of solving the matrix equation Ax=b. This course provides you with arms to tackle the various computational problems you face in solving the equation.

Course Content:

Basic iterative methods for solutions of linear systems and their rates of convergence. Generalized conjugate gradient, Krylov space and Lanczos methods. Iterative methods for symmetric, non-symmetric and generalized eigenvalue problems. Singular value decompositions. Fast computations for structured matrices. Polynomial matrix computations. Perturbation bounds for eigenvalues.

Pre requisites:

MA106 Linear Algebra. Having a background will help a lot. But you can manage with some difficulty if you have not done Linear Algebra before.

Feedback on Lectures:

Lectures were one of the most important things in understanding the course. A few concepts in the textbook were a bit involved and difficult to understand in the first reading itself. Attending lectures would mean an easier understanding of the concepts when you read the book. The professor would solve each and every problem without hesitation. Infact, the professor would allot time in the beginning for doubts of previous lectures. In short, missing a lecture would mean a lot of hours to be spent on understanding that particular topic.

Feedback on Exams:

If you were expecting them to be easy, then you are in for a ride. Inspite of having a firm grip on the concepts, scoring in the exams was difficult. The papers were lengthy and tough. But since it is relative, you can be safe if you have studied well because there will be a few questions which you can crack pretty easily. Also, a few assignments are given which usually have a very less weightage.

Grading:

The professor is known to be a bit tough on grading. Please have a look on the ASC portal for the same.

Attendance:

No policy as such. But not attending implies a weak performance in the exams.

Books:

David Watkins Fundamentals of Matrix Computations

Future direction:

Everywhere where you have to solve the equation Ax=b. Signal processing and image processing.

Reviewed by:

Sravan Patchala (sravps7@gmail.com)