An Introduction to Reinforcement Learning

Welcome to the study of reinforcement learning! This textbook accompanies the undergraduate course CS 1840/STAT 184 taught at Harvard. It is intended to be an approachable yet rigorous introduction to this active subfield of machine learning. This book assumes the same prerequisites as the course: You should be familiar with multivariable calculus, linear algebra, and probability. For Harvard undergraduates, this is fulfilled by Math 21a, Math 21b, and Stat 110, or their equivalents. Stat 111 is strongly recommended but not required. Specifically, we will assume that you know the following topics. The italicized terms have brief re-introductions in the text: You should also be familiar with basic programming concepts such as variables, functions, loops, etc. Pseudocode listings will be provided for certain algorithms. The course will progress through the following units: 1  Introduction presents motivation for the RL problem and compares RL to other fields of machine learning. 2  Marko

saved by

• Grace S