6.7950 Fall 2022 (formerly 6.246)      

Reinforcement Learning: Foundations and Methods

 

Note: Not yet updated for Fall 2022.

Lecture scribing instruction (extra credit):

Lecture Date Subject Optional Readings Deadlines
PART 1: Dynamic Programming
1T 02/16Reinforcement learning overview [slides] [notes]DPOC vol 1, Ch 1-2HW0 out (not due)
2R 02/18Dynamic programming [slides]DPOC vol 1, Ch 2-3.4HW0 (post solutions)
HW1 out (1.5 weeks)
R1F 02/19Dynamic programming (finite horizon)
3T 02/23Markov Decision Processes [slides][notes]DPOC vol 1, Ch 5.1, 5.4;
DPOC vol 2, Ch 1.1-1.5
4R 02/25Markov Decision Processes
R2F 02/26Dynamic programming (stochastic, infinite horizon)
5T 03/02Undiscounted infinite horizon dynamic programming [slides] [notes] DPOC vol 1, Ch 5.2-5.3,5.5;
DPOC vol 2, Ch 3.1-3.2
HW1 due
HW2 out (2 weeks)
6R 03/04Undiscounted infinite horizon dynamic programming [notes]
R3F 03/05Infinite horizon dynamic programming
T 03/09NO CLASS (Monday Schedule)
PART 2: Approximate Dynamic Programming
7R 03/11Model-free reinforcement learning [slides]
R4F 03/12Infinite horizon dynamic programming
8T 03/16Stochastic approximation [slides][notes]NDP, Ch 3.2, Ch 4.1, 4.3, 5HW2 due
HW3 out
(2 weeks)
Sign-up for student lectures out
9R 03/18Representation approximation [slides][notes]NDP, Ch 3.1, Ch 4.2, Ch 6.1-6.3
R5F 03/19Stochastic & representation approximationNDP, Ch 5
T 03/23NO CLASS (Student holiday)Sign-up for student lectures due
10R 03/25Representation approximation [notes]
R6F 03/26Representation approximation
11T 03/30Policy space methods (Part 1) [slides][notes]
12R 04/01Policy space methods (Part 2)HW3 due
HW4 out
(1.5 weeks)
R7F 04/02Policy space methods
13T 04/06Intro to Multi-Arm Bandits [slides] Project Proposal due
14R 04/08Application of RL [slides]
F 04/09NO RECITATION
15T 04/13State Abstraction (Student Lecture) [slides] [notes]HW4 due
16R 04/15Exploration vs Exploitation (Student Lecture) [slides] [notes]Project proposal feedback back to students
R8F 04/16Quiz review
T 04/20NO CLASS (Student holiday)
17R 04/22QuizQuiz
HW5 out (2 weeks)
R9F 04/23Implementation
18T 04/27Transfer and Curriculum Learning (Student Lecture) [slides] [notes]
19R 04/29Off-policy RL (Student Lecture) [slides][notes]
20T 05/04Multi-agent Deep RL (Student Lecture) [slides] [notes]
21R 05/06Safe RL (Student Lecture) [slides][notes]HW5 due
22T 05/11Cooperation and Competition (Student Lecture) [slides][notes]
23R 05/13Model-based RL (Student Lecture) [slides][notes]
24T 05/18Project presentationsProject presentation slides (before class)
25R 05/20Project presentationsProject presentation slides (before class)
Final project reports (5pm)