Planning and Using Models in Reinforcement Learning

Summary

This course-lecture summary introduces the role of planning and learned models in reinforcement learning. It identifies Dyna and Monte Carlo tree search as examples of algorithms that use models, and notes that the lecture is presented by research engineer Matteo Hessel. The subject is relevant to quantitative researchers exploring how an agent can learn a representation of its environment and use that representation to guide decisions.

The supplied text is a brief catalog description rather than lecture notes. It gives no explanation of how Dyna combines learning from experience with planning, how Monte Carlo tree search explores possible future outcomes, or how either method is implemented or evaluated. It also contains no financial-market application, experiments, or performance evidence. Readers would need the underlying lecture or other technical material to learn the algorithms in detail and assess when they may be useful for trading research.

Key ideas

• The lecture covers planning and the use of learned models in reinforcement learning.
• Dyna and Monte Carlo tree search are named as example model-based algorithms.
• The description gives the lecture topic and presenter but not the algorithms’ mechanics.
• No trading application, experimental evidence, or performance results are included in the supplied text.

Tags

This summary was written by Stratmill's research agent from the original; it is not a copy of the source.