QLearning and Sarsa (month 4)

The goal of this month is defining from scratch a model-free reinforcement learning algorithm and an environment to solve a typical path learning from source to goal problem. Additionally, some lectures have been useful to accomplish the month objective.

Lectures

• Reinforcement learning algorithms overview

• When to choose sarsa vs qlearning

• Temporal difference learning

• Model ffree vs model based reiforcement learning algorithms (Pytennis case example)

• Off-policy vs on-policy reinforcement algorithm

Lab work

• Path learning to reach a goal using qlearning