COMPSCI 188 - 2018-09-25 - Reinforcement Learning Part 1/2

Рет қаралды 32,576

Күн бұрын

COMPSCI 188, LEC 001 - Fall 2018
COMPSCI 188, LEC 001 - Pieter Abbeel, Daniel Klein
Copyright @2018 UC Regents; all rights reserved
"Slides (from 2018): inst.eecs.berkeley.edu/~cs188...
Latest website: inst.eecs.berkeley.edu/~cs188
More resources: ai.berkeley.edu
00:00 Setup [no content]
02:03 Announcements [outdated]
05:13 RL Introduction
07:15 RL Applications
15:26 RL Definition
18:40 Model-Based Learning
28:15 Model-Based vs. Model-Free Estimation
34:18 Passive RL
35:51 Direct Evaluation
41:40 Sample-Based Policy Evaluation?
45:47 Temporal Difference Learning
50:26 TD Learning: Example
53:33 Break [no content]
55:55 Problems with TD Learning
1:00:32 Active RL
1:02:01 Q-Value Iteration
1:05:50 Q-Learning
1:16:43 Q-Learning: Crawler Bot Demo
1:18:38 Q-Learning Properties
1:19:53 End [no content]"

Пікірлер: 10

@RahulKumar-tc3kq 3 жыл бұрын

not only these lectures are good, but also the cartoons in these lecture slides are awesome.

@mystmuffin3600 2 жыл бұрын

35:00 😂

@charlesz88 4 жыл бұрын

that due yawning was hilarious 23:49

@quiteSimple24 5 жыл бұрын

starts 2:04 lecture starts 5:14

@chococ7651 2 жыл бұрын

great lecture

@kailinliang2408 5 жыл бұрын

neuq的小伙伴们你们好啊

@user54246 4 жыл бұрын

end of break time 55:56

@hangchen 4 жыл бұрын

So solving MDPs when all the properties(S, a, T, R) are known is considered as offline learning and RL is considered as online learning? Are all RL online learning? Is RL in simulation also online learning? Thanks!