[cs231n] Lecture 14 | Deep Reinforcement Learning

Lecture 14 | Deep Reinforcement Learning

  • What is Reinforcement Learning?
  • Markov Decision Processes
  • Q-Learning 
  • Policy Gradients


Cart-Pole Problem

Robot Locomotion

Atari Games

GO

Markov Decision Process

A simple MDP: Grid World

Definitions: Value function and Q-value function

Bellman equation

Solving for the optimal policy

Solving for the optimal policy: Q-learning

Case Study: Playing Atari Games

Q-network Architecture

Training the Q-network: Loss function (from before)

Training the Q-network: Experience Replay

Putting it together: Deep Q-Learning with Experience Replay


Policy Gradients

REINFORCE algorithm

Intuition

Variance reduction

Variance reduction: Baseline

How to choose the baseline?

Actor-Critic Algorithm

REINFORCE in action: Recurrent Attention Model (RAM)

  • 0
    点赞
  • 1
    收藏
    觉得还不错? 一键收藏
  • 0
    评论

“相关推荐”对你有帮助么?

  • 非常没帮助
  • 没帮助
  • 一般
  • 有帮助
  • 非常有帮助
提交
评论
添加红包

请填写红包祝福语或标题

红包个数最小为10个

红包金额最低5元

当前余额3.43前往充值 >
需支付:10.00
成就一亿技术人!
领取后你会自动成为博主和红包主的粉丝 规则
hope_wisdom
发出的红包
实付
使用余额支付
点击重新获取
扫码支付
钱包余额 0

抵扣说明:

1.余额是钱包充值的虚拟货币,按照1:1的比例进行支付金额的抵扣。
2.余额无法直接购买下载,可以购买VIP、付费专栏及课程。

余额充值