Reducing trials by thinning-out in skill discovery

Hayato Kobayashi; Kohei Hatano; Akira Ishino; Ayumi Shinohara

doi:10.1007/978-3-540-75488-6_13

Reducing trials by thinning-out in skill discovery

Hayato Kobayashi, Kohei Hatano, Akira Ishino, Ayumi Shinohara

Division for Theoretical Natural Science

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution

1 Citation (Scopus)

Abstract

In this paper, we propose a new concept, thinning-out, for reducing the number of trials in skill discovery. Thinning-out means to skip over such trials that are unlikely to improve discovering results, in the same way as "pruning" in a search tree. We show that our thinning-out technique significantly reduces the number of trials. In addition, we apply thinning-out to the discovery of good physical motions by legged robots in a simulation environment. By using thinning-out, our virtual robots can discover sophisticated motions that is much different from the initial motion in a reasonable amount of trials.

Original language	English
Title of host publication	Discovery Science - 10th International Conference, DS 2007, Proceedings
Publisher	Springer Verlag
Pages	127-138
Number of pages	12
ISBN (Print)	9783540754879
DOIs	https://doi.org/10.1007/978-3-540-75488-6_13
Publication status	Published - 2007
Event	10th International Conference on Discovery Science, DS 2007 - Sendai, Japan Duration: Oct 1 2007 → Oct 4 2007

Publication series

Name	Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Volume	4755 LNAI
ISSN (Print)	0302-9743
ISSN (Electronic)	1611-3349

Other

Other	10th International Conference on Discovery Science, DS 2007
Country/Territory	Japan
City	Sendai
Period	10/1/07 → 10/4/07

All Science Journal Classification (ASJC) codes

Theoretical Computer Science
Computer Science(all)

Access to Document

10.1007/978-3-540-75488-6_13

Cite this

Kobayashi, H., Hatano, K., Ishino, A., & Shinohara, A. (2007). Reducing trials by thinning-out in skill discovery. In Discovery Science - 10th International Conference, DS 2007, Proceedings (pp. 127-138). (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 4755 LNAI). Springer Verlag. https://doi.org/10.1007/978-3-540-75488-6_13

Reducing trials by thinning-out in skill discovery. / Kobayashi, Hayato; Hatano, Kohei; Ishino, Akira et al.

Discovery Science - 10th International Conference, DS 2007, Proceedings. Springer Verlag, 2007. p. 127-138 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 4755 LNAI).

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution

Kobayashi, H, Hatano, K, Ishino, A & Shinohara, A 2007, Reducing trials by thinning-out in skill discovery. in Discovery Science - 10th International Conference, DS 2007, Proceedings. Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), vol. 4755 LNAI, Springer Verlag, pp. 127-138, 10th International Conference on Discovery Science, DS 2007, Sendai, Japan, 10/1/07. https://doi.org/10.1007/978-3-540-75488-6_13

Kobayashi H, Hatano K, Ishino A, Shinohara A. Reducing trials by thinning-out in skill discovery. In Discovery Science - 10th International Conference, DS 2007, Proceedings. Springer Verlag. 2007. p. 127-138. (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)). doi: 10.1007/978-3-540-75488-6_13

@inproceedings{e46670a84dc54adea401db3eae63effb,

title = "Reducing trials by thinning-out in skill discovery",

abstract = "In this paper, we propose a new concept, thinning-out, for reducing the number of trials in skill discovery. Thinning-out means to skip over such trials that are unlikely to improve discovering results, in the same way as {"}pruning{"} in a search tree. We show that our thinning-out technique significantly reduces the number of trials. In addition, we apply thinning-out to the discovery of good physical motions by legged robots in a simulation environment. By using thinning-out, our virtual robots can discover sophisticated motions that is much different from the initial motion in a reasonable amount of trials.",

author = "Hayato Kobayashi and Kohei Hatano and Akira Ishino and Ayumi Shinohara",

year = "2007",

doi = "10.1007/978-3-540-75488-6_13",

language = "English",

isbn = "9783540754879",

series = "Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)",

publisher = "Springer Verlag",

pages = "127--138",

booktitle = "Discovery Science - 10th International Conference, DS 2007, Proceedings",

address = "Germany",

note = "10th International Conference on Discovery Science, DS 2007 ; Conference date: 01-10-2007 Through 04-10-2007",

}

TY - GEN

T1 - Reducing trials by thinning-out in skill discovery

AU - Kobayashi, Hayato

AU - Hatano, Kohei

AU - Ishino, Akira

AU - Shinohara, Ayumi

PY - 2007

Y1 - 2007

N2 - In this paper, we propose a new concept, thinning-out, for reducing the number of trials in skill discovery. Thinning-out means to skip over such trials that are unlikely to improve discovering results, in the same way as "pruning" in a search tree. We show that our thinning-out technique significantly reduces the number of trials. In addition, we apply thinning-out to the discovery of good physical motions by legged robots in a simulation environment. By using thinning-out, our virtual robots can discover sophisticated motions that is much different from the initial motion in a reasonable amount of trials.

AB - In this paper, we propose a new concept, thinning-out, for reducing the number of trials in skill discovery. Thinning-out means to skip over such trials that are unlikely to improve discovering results, in the same way as "pruning" in a search tree. We show that our thinning-out technique significantly reduces the number of trials. In addition, we apply thinning-out to the discovery of good physical motions by legged robots in a simulation environment. By using thinning-out, our virtual robots can discover sophisticated motions that is much different from the initial motion in a reasonable amount of trials.

UR - http://www.scopus.com/inward/record.url?scp=38149073337&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=38149073337&partnerID=8YFLogxK

U2 - 10.1007/978-3-540-75488-6_13

DO - 10.1007/978-3-540-75488-6_13

M3 - Conference contribution

AN - SCOPUS:38149073337

SN - 9783540754879

T3 - Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)

SP - 127

EP - 138

BT - Discovery Science - 10th International Conference, DS 2007, Proceedings

PB - Springer Verlag

T2 - 10th International Conference on Discovery Science, DS 2007

Y2 - 1 October 2007 through 4 October 2007

ER -

Reducing trials by thinning-out in skill discovery

Abstract

Publication series

Other

All Science Journal Classification (ASJC) codes

Access to Document

Other files and links

Fingerprint

Cite this