Reinforcement Learning 8: Policy gradient methods

Описание к видео Reinforcement Learning 8: Policy gradient methods

Slides: https://cwkx.github.io/data/teaching/...
Code: https://github.com/higgsfield/RL-Adve...
Theory: https://lilianweng.github.io/lil-log/...
Twitter:   / cwkx  
Next video:    • Reinforcement Learning Lectures  

Policy-based methods
- definition
- characteristics
- deterministic vs stochastic policies
Policy gradients
- gradient-based estimator
- Monte Carlo REINFORCE
Actor-critic methods
- definition
- algorithm
- extensions

#policygradients #actorcritic #reinforcementlearning #REINFORCE #montecarlo

Комментарии

Информация по комментариям в разработке