Learning from observation and from practice using behavioral primitives

Darrin Bentivegna; Gordon Cheng; Christopher Atkeson

doi:10.1007/11008941_59

Learning from observation and from practice using behavioral primitives

Darrin Bentivegna, Gordon Cheng, Christopher Atkeson

Research output: Contribution to journal › Article › peer-review

4 Scopus citations

Abstract

We describe a memory-based approach to learning how to select and provide subgoals for behavioral primitives, given an existing library of primitives. We demonstrate both learning from observation and learning from practice on a marble maze task, Labyrinth.

Original language	English
Pages (from-to)	551-560
Number of pages	10
Journal	Springer Tracts in Advanced Robotics
Volume	15
DOIs	https://doi.org/10.1007/11008941_59
State	Published - 2005
Externally published	Yes

Access to Document

10.1007/11008941_59

Cite this

@article{6ee83b599ecd45afa1bebd78737490ac,

title = "Learning from observation and from practice using behavioral primitives",

abstract = "We describe a memory-based approach to learning how to select and provide subgoals for behavioral primitives, given an existing library of primitives. We demonstrate both learning from observation and learning from practice on a marble maze task, Labyrinth.",

author = "Darrin Bentivegna and Gordon Cheng and Christopher Atkeson",

year = "2005",

doi = "10.1007/11008941_59",

language = "English",

volume = "15",

pages = "551--560",

journal = "Springer Tracts in Advanced Robotics",

issn = "1610-7438",

publisher = "Springer Verlag",

}

TY - JOUR

T1 - Learning from observation and from practice using behavioral primitives

AU - Bentivegna, Darrin

AU - Cheng, Gordon

AU - Atkeson, Christopher

PY - 2005

Y1 - 2005

N2 - We describe a memory-based approach to learning how to select and provide subgoals for behavioral primitives, given an existing library of primitives. We demonstrate both learning from observation and learning from practice on a marble maze task, Labyrinth.

AB - We describe a memory-based approach to learning how to select and provide subgoals for behavioral primitives, given an existing library of primitives. We demonstrate both learning from observation and learning from practice on a marble maze task, Labyrinth.

UR - http://www.scopus.com/inward/record.url?scp=84885081832&partnerID=8YFLogxK

U2 - 10.1007/11008941_59

DO - 10.1007/11008941_59

M3 - Article

AN - SCOPUS:84885081832

SN - 1610-7438

VL - 15

SP - 551

EP - 560

JO - Springer Tracts in Advanced Robotics

JF - Springer Tracts in Advanced Robotics

ER -

Learning from observation and from practice using behavioral primitives

Abstract

Access to Document

Other files and links

Fingerprint

Cite this