What is reinforcement learning and when is it useful? How can it be implemented and applied in Python? This talk will attempt to answer these questions.

1) Introduction & Concepts

### Definition of Reinforcement Learning ### Relevant Concepts * Markov decision process * state space * action space * transition probabilities * state rewards * policy ### Using Data * estimating transition probabilities and rewards from data * identifying the optimal policy given estimated transition probabilities

2) Implementation in Python

• Code sample: estimating the state-transition-probability tensor
• Code sample: estimating rewards
• Code sample: iteratively solving for the optimal policy

3) Demo

• example problem
• demo use of a reinforcement learning to solve in Python