Sign inSign up
Agnostic KWIK learning and efficient approximate reinforcement learning thumbnail
Pause
Mute
Subtitles
Playback speed
0.25
0.5
0.75
1
1.25
1.5
1.75
2
Full screen

Agnostic KWIK learning and efficient approximate reinforcement learning

Published on Aug 02, 20114024 Views

Csaba Szepesvári

A popular approach in reinforcement learning is to use a model-based algorithm, i.e., an algorithm that utilizes a model learner to learn an approximate model to the environment. It has been shown suc

COLT 2011 - Budapest

Related categories

Reinforcement Learning

Chapter list

Agnostic KWIK learning and eficient approximate reinforcement learning00:00
Outline00:28
Reinforcement learning - 101:26
Reinforcement learning - 201:49
Reinforcement learning - 301:59
Eficient RL algorithms02:57
The "Rmax-construction": A general scheme for eficient RL - 103:54
The "Rmax-construction": A general scheme for eficient RL - 205:57
The "Knows what it knows" (KWIK) framework - 106:51
The "Knows what it knows" (KWIK) framework - 207:21
The "Knows what it knows" (KWIK) framework - 307:36
The "Knows what it knows" (KWIK) framework - 407:45
The "Knows what it knows" (KWIK) framework - 508:03
The "Knows what it knows" (KWIK) framework - 608:10
The "Knows what it knows" (KWIK) framework - 708:16
The Rmax construction with a KWIK learner - 108:45
The Rmax construction with a KWIK learner - 210:09
The KWIK-Rmax theorem - 111:27
The KWIK-Rmax theorem - 212:11
The need for agnostic learning12:21
Agnostic KWIK learning12:47
Agnostic KWIK learning: prediction accuracy - 113:35
Agnostic KWIK learning: prediction accuracy - 213:36
Agnostic KWIK learning: prediction accuracy - 313:36
Problems and problem classes13:51
Agnostic KWIK learner - 113:53
Agnostic KWIK learner - 213:56
Agnostic KWIK-Rmax theorem14:03
The agnostic KWIK-Rmax theorem justi es the agnostic KWIK framework! - 114:36
The agnostic KWIK-Rmax theorem justi es the agnostic KWIK framework! - 214:41
Finite hypothesis class H, deterministic case - 114:54
Finite hypothesis class H, deterministic case - 215:05
Finite hypothesis class H, deterministic case - 315:08
Finite hypothesis class H, deterministic case - 415:24
Finite hypothesis class H, deterministic case - 515:30
Finite hypothesis class H, deterministic case - 615:33
Finite hypothesis class H, deterministic case - 715:35
Finite hypothesis class H, deterministic case - 815:37
Finite hypothesis class H, deterministic case - 915:38
A sample run of the agnostic KWIK learner - 115:57
A sample run of the agnostic KWIK learner - 215:59
A sample run of the agnostic KWIK learner - 315:59
A sample run of the agnostic KWIK learner - 416:00
A sample run of the agnostic KWIK learner - 516:00
A sample run of the agnostic KWIK learner - 616:01
A sample run of the agnostic KWIK learner - 716:15
A sample run of the agnostic KWIK learner - 816:17
A sample run of the agnostic KWIK learner - 916:18
A sample run of the agnostic KWIK learner - 1016:18
A sample run of the agnostic KWIK learner - 1116:18
A sample run of the agnostic KWIK learner - 1216:19
A sample run of the agnostic KWIK learner - 1316:19
Finite hypothesis class H, noisy problems - 116:20
Finite hypothesis class H, noisy problems - 216:22
Finite hypothesis class H, noisy problems - 316:41
Finite hypothesis class H, noisy problems - 416:47
Finite hypothesis class H, noisy problems - 517:01
Finite hypothesis class H, noisy problems - 617:01
Finite hypothesis class H, noisy problems - 717:01
Finite hypothesis class H, noisy problems - 817:02
Finite hypothesis class H, noisy problems - 917:02
Finite hypothesis class H, noisy problems - 1017:02
Finite hypothesis class H, noisy problems - 1117:03
Finite hypothesis class H, noisy problems - 1217:03
Finite hypothesis class H, noisy problems - 1317:04
Table of learning complexities17:06
Summary - 118:26
Summary - 219:22