Learning path
Reinforcement Learning Basics
Agents, rewards and exploration, and Q-learning you can watch learn, with animated films and runnable Python. Part of the Data Scientist premium track.
-
1
Reinforcement Learning: Agents, Rewards and Environments
Learning by trial and error: states, actions, rewards, policies, the explore-exploit trade-off with a multi-armed bandit, and where RL is used, from games to RLHF for chatbots.
-
2
Q-Learning on a Grid World
Train an agent to find the exit of a maze with Q-learning: the Q-table, the update rule, epsilon-greedy exploration, and how Deep Q-Networks scale the same idea, all…