Topics:
- Study applications of Reinforcement Learning for stock trading
- Discuss various problems in Quantitative Trading that amount to Reinforcement Learning tasks
- Study such problems as optimal portfolio execution, dynamic portfolio management, and index tracking
- Develop a simple portfolio model that allows us to address all these problems in the same modeling framework
- Explain a RL approach to such problem, that learns optimal trading or execution policy directly from data made of states, actions, and rewards
- Analyze an Inverse Reinforcement Learning setting for these problems, when we don't observe rewards, and will see why it may be more useful than a Reinforcement Learning setting, in many problems of practical interest.
- Study how Inverse Reinforcement Learning works in this setting, and how it can be applied to learn a reward function, also called a utility function, of an investor.
- Apply Inverse Reinforcement Learning to all investors simultaneously, and show how we can use the same to learn market-optimal trading strategies
This is a translated note for the coursera course (linked below).
https://www.coursera.org/learn/reinforcement-learning-in-finance/
