Learn the fundamentals of Reinforcement Learning (RL), including Markov Decision Processes, Q-learning, SARSA, and an introduction to Deep Q-Networks. This course provides theoretical understanding and practical implementation skills for building agents that learn optimal behaviors through interaction with an environment. Suitable for engineers and developers with existing knowledge of machine learning principles.
Prerequisites Machine Learning & Python
Level:
RL Fundamentals
Understand the core components of RL: agents, environments, states, actions, rewards, policies.
Markov Decision Processes
Formulate problems using MDPs and understand their properties.
Value Functions & Bellman Equations
Grasp the concepts of state-value and action-value functions and derive Bellman equations.
Model-Free Learning
Implement and differentiate between Monte Carlo and Temporal Difference learning methods (Q-learning, SARSA).
Function Approximation
Apply function approximation techniques in RL for large state spaces.
Deep Q-Networks (DQN)
Understand the basic principles behind DQN and its components like experience replay.
RL Implementation
Build and train simple RL agents using Python and relevant libraries.