Reinforcement Learning for Business Applications

Course Objectives

  1. To introduce theory of Markov decision problems and reinforcement learning
  2. To introduce different reinforcement learning techniques
  3. To implement different reinforcement learning techniques to various problems
  4. To introduce performance evaluation of reinforcement learning techniques

Topics

  • Reinforcement Learning Foundations
  • Multi-Armed Bandit
  • Markov Decision Processes
  • Value Iteration – Policy Iteration
  • Monte Carlo Methods
  • Q-Learning, SARSA
  • Eligibility Traces
  • Exploration vs. Exploitation
  • Function Approximation: Stochastic-gradient, Semi-gradient TD Update, Least-squares TD
  • Value-based Deep RL: Q-network
  • Policy-based Deep RL: REINFORCE