ROB 537
Fall 2017
4 credits

Learning Based Control
MW 10:00-11:15
Batcheler 250

Instructor: Kagan Tumer
Office: Graf 314
Office hours: MW: 11:30-12:15
Help Sessions: Shaw Khadka
Location: Covell 140
Time: M: 3:00-4:00 and by appointment


Recommended books:

Grading :

Homework 20%
Midterm Exam 40%
Research Project     40%

Learning Outcomes: By the end of the course students will be able to:

Students with Disabilities: Accommodations are collaborative efforts between students, faculty and Services for Students with Disabilities (SSD). Students with accommodations approved through SSD are responsible for contacting the faculty member in charge of the course prior to or during the first week of the term to discuss accommodations. Students who believe they are eligible for accommodations but who have not yet obtained approval through SSD should contact SSD immediately at 737-4098.


Week Date Lecture Title Homework Project
  9/20 Course Introduction    
1 9/25 Neural Network Basics  

Groups Finalized

9/27 Feed Forward Neural Networks    
2 10/2 Search and Optimization HW 1 due  
10/4 Evolutionary Algorithms and MOO    
3 10/9 Deep Learning  

Project Topic due

10/11 Project Discussion    
4 10/16 Neural Networks for Control HW 2 due  
10/18 Reinforcement Learning    
5 10/23 Temporal Difference Learning  

Project Background due

10/25 Policy Gradient & Transfer Learning    
6 10/30 Recurrent Neural Networks HW 3 due  
11/1 Planning    
7 11/6 MIDTERM EXAM    
11/8 Project Discussion    
8 11/13 State Estimation and Bayes Filters HW 4 due  
11/15 Ethics, Policy, Law    
9 11/20 Reward Estimation  

Preliminary paper due

11/22 Project Update    
10 11/27 Final Presentations - Rog 226    
11 12/4    

Final paper due

Academic Dishonesty: You are permitted, and to a great extent encouraged, to work with others on homework sets. However, there is an obvious difference between constructive discussion of a particular problem and copying. Acts of academic dishonesty will not be tolerated and will be handled according to university policy. (See for details.)