Reinforcement Learning

Preface

Reinforcement learning (RL) is a machine learning (ML) paradigm that is capable of optimizing sequential decisions. RL is interesting because it mimics how we, as humans, learn. We are instinctively capable of learning strategies that help us master complex tasks like riding a bike or taking a mathematics exam. RL attempts to copy this process by interacting with the environment to learn strategies.

Recently, businesses have been applying ML algorithms to make one-shot decisions. These are trained upon data to make the best decision at the time. But often, the right decision at the time may not be the best decision in the long term. Yes, that full tub of ice cream will make you happy in the short term, but you’ll have to do more exercise next week. Similarly, click-bait recommendations might have the highest click-through rates, but in the long term these articles feel like a scam and hurt long-term engagement or retention.

RL is exciting because it is possible to learn long-term strategies and apply them to complex industrial problems. Businesses and practitioners alike can use goals that directly relate to the business like profit, number of users, and retention, not technical evaluation metrics like accuracy or F1-score. Put simply, many challenges depend on sequential decision making. ML is not designed to solve these problems, RL is.

Objective

I wrote this book because I have read about so many amazing examples of using RL to solve seemingly impossible tasks. But ...