COMPSCI 188 - 2018-09-20 - Markov Decision Processes (MDPs) Part 2/2
Рет қаралды 29,454
Күн бұрынCOMPSCI 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]
03:22 Contest Results [outdated]
09:22 Review: MDPs
20:26 The Bellman Equations
25:41 Value Iteration
29:21 Convergence of Value Iteration
33:15 Policy Evaluation
39:02 Policy Evaluation: Example
41:14 Policy Evaluation: Computation
44:39 Policy Extraction
50:41 Break [no content]
53:28 Problems with Value Iteration
56:37 Policy Iteration
1:02:03 Policy Iteration: Q&A, Summary
1:06:26 Reinforcement Learning: Slots Demo
1:13:04 Reinforcement Learning Preview
1:16:40 End [no content]"
@skittles6486 4 жыл бұрын
Awesome explanation. Now I clearly understood why UC Berkeley is so special, not only for their tag, but mainly for the quality. This video is nothing but a proof
@ja6091 5 жыл бұрын
Lecture actually starts at 9:25
@sanjanachopra2100 4 жыл бұрын
Topics: Markov Decision Process II, Bellman equations, Convergence, policy methods, Policy evaluation, Policy iteration.
@user-or7ji5hv8y 4 жыл бұрын
why can't we see a demo of policy iteration like it was shown with grid world for value iteration? is it harder to visually demo?
@ShoniaVika 4 жыл бұрын
53:32 break ends
@schlutzzz 5 жыл бұрын
Start 3:23
Mutual Information
Рет қаралды 40 М.
Webcast Departmental
Рет қаралды 31 М.