CMSC 828R Machine Learning: Spectral Methods and Reinforcement Learning

 


Fall/2017

CSI 2120 (Bldg. # 406, map)

Tuesday Thursday 12:30 pm – 1:45 pm

 

Instructor:                   Furong Huang <furongh@cs.umd.edu>

Office Hours:              Thursday 2:00 pm – 3:00 pm, 3209 A.V. Williams

Grading:                      3 credits, Regular

TA:                              Chengxi Ye <yechengxi@gmail.com>

Piazza:                         Website (access code provided in class)

 

 

 

 

This course will focus on advanced machine learning algorithms that have provable guarantees. In this course, we will focus on surveying various fundamental algorithms whose performance can be rigorously analyzed.  Two major fields will be introduced: spectral methods and reinforcement learning.  We will cover spectral methods related topics such as nonnegative matrix factorization, tensor decomposition, and learning mixture models. Introductory materials in reinforcement learning and spectral methods in reinforcement learning will also be covered.

 

 

Syllabus:

 

Detailed Syllabus:

 

 


[Back] Back to the Department of Computer Science Class Pages


Web Accessibility