Family bootstrapping: A genetic transfer learning approach for onsetting the evolution for a set of related robotic tasks

Amiram Moshaiov; Amir Tal

doi:10.1109/CEC.2014.6900571

Family bootstrapping: A genetic transfer learning approach for onsetting the evolution for a set of related robotic tasks

Amiram Moshaiov^*
, Amir Tal

^*Corresponding author for this work

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

13 Scopus citations

Abstract

Studies on the bootstrap problem in evolutionary robotics help lifting the barrier from the way to evolve robots for complex tasks. It remains an open question, though, how to reduce the need for designer knowledge when devising a bootstrapping approach for any particular complex task. Transfer learning may help reducing this need and support the evolution of solutions to complex tasks, through task relatedness. Relying on the commonalities of similar tasks, we introduce a new concept of Family Bootstrapping (FB). FB refers to the creation of biased ancestors that are expected to onset the evolution of 'a family' of solutions not just for one task, but for a set of related robot tasks. A general FB paradigm is outlined and the unique potential of the proposed concept is discussed. To highlight the validity of the FB concept, a simple demonstration case, concerning the evolution of neuro-controllers for a set of robot navigation tasks, is provided. The paper is concluded with some suggestions for future research.

Original language	English
Title of host publication	Proceedings of the 2014 IEEE Congress on Evolutionary Computation, CEC 2014
Publisher	Institute of Electrical and Electronics Engineers Inc.
Pages	2801-2808
Number of pages	8
ISBN (Electronic)	9781479914883
DOIs	https://doi.org/10.1109/CEC.2014.6900571
State	Published - 16 Sep 2014
Externally published	Yes
Event	2014 IEEE Congress on Evolutionary Computation, CEC 2014 - Beijing, China Duration: 6 Jul 2014 → 11 Jul 2014

Publication series

Name	Proceedings of the 2014 IEEE Congress on Evolutionary Computation, CEC 2014

Conference

Conference	2014 IEEE Congress on Evolutionary Computation, CEC 2014
Country/Territory	China
City	Beijing
Period	6/07/14 → 11/07/14

Bibliographical note

Publisher Copyright:
© 2014 IEEE.

Access to Document

10.1109/CEC.2014.6900571

Cite this

Moshaiov, A., & Tal, A. (2014). Family bootstrapping: A genetic transfer learning approach for onsetting the evolution for a set of related robotic tasks. In Proceedings of the 2014 IEEE Congress on Evolutionary Computation, CEC 2014 (pp. 2801-2808). Article 6900571 (Proceedings of the 2014 IEEE Congress on Evolutionary Computation, CEC 2014). Institute of Electrical and Electronics Engineers Inc.. https://doi.org/10.1109/CEC.2014.6900571

Moshaiov, Amiram ; Tal, Amir. / Family bootstrapping : A genetic transfer learning approach for onsetting the evolution for a set of related robotic tasks. Proceedings of the 2014 IEEE Congress on Evolutionary Computation, CEC 2014. Institute of Electrical and Electronics Engineers Inc., 2014. pp. 2801-2808 (Proceedings of the 2014 IEEE Congress on Evolutionary Computation, CEC 2014).

@inproceedings{aae5152bc7d048ef85a4f7eae7cfeea6,

title = "Family bootstrapping: A genetic transfer learning approach for onsetting the evolution for a set of related robotic tasks",

abstract = "Studies on the bootstrap problem in evolutionary robotics help lifting the barrier from the way to evolve robots for complex tasks. It remains an open question, though, how to reduce the need for designer knowledge when devising a bootstrapping approach for any particular complex task. Transfer learning may help reducing this need and support the evolution of solutions to complex tasks, through task relatedness. Relying on the commonalities of similar tasks, we introduce a new concept of Family Bootstrapping (FB). FB refers to the creation of biased ancestors that are expected to onset the evolution of 'a family' of solutions not just for one task, but for a set of related robot tasks. A general FB paradigm is outlined and the unique potential of the proposed concept is discussed. To highlight the validity of the FB concept, a simple demonstration case, concerning the evolution of neuro-controllers for a set of robot navigation tasks, is provided. The paper is concluded with some suggestions for future research.",

author = "Amiram Moshaiov and Amir Tal",

note = "Publisher Copyright: {\textcopyright} 2014 IEEE.; 2014 IEEE Congress on Evolutionary Computation, CEC 2014 ; Conference date: 06-07-2014 Through 11-07-2014",

year = "2014",

month = sep,

day = "16",

doi = "10.1109/CEC.2014.6900571",

language = "אנגלית",

series = "Proceedings of the 2014 IEEE Congress on Evolutionary Computation, CEC 2014",

publisher = "Institute of Electrical and Electronics Engineers Inc.",

pages = "2801--2808",

booktitle = "Proceedings of the 2014 IEEE Congress on Evolutionary Computation, CEC 2014",

address = "ארצות הברית",

}

Moshaiov, A & Tal, A 2014, Family bootstrapping: A genetic transfer learning approach for onsetting the evolution for a set of related robotic tasks. in Proceedings of the 2014 IEEE Congress on Evolutionary Computation, CEC 2014., 6900571, Proceedings of the 2014 IEEE Congress on Evolutionary Computation, CEC 2014, Institute of Electrical and Electronics Engineers Inc., pp. 2801-2808, 2014 IEEE Congress on Evolutionary Computation, CEC 2014, Beijing, China, 6/07/14. https://doi.org/10.1109/CEC.2014.6900571

Family bootstrapping: A genetic transfer learning approach for onsetting the evolution for a set of related robotic tasks. / Moshaiov, Amiram; Tal, Amir.
Proceedings of the 2014 IEEE Congress on Evolutionary Computation, CEC 2014. Institute of Electrical and Electronics Engineers Inc., 2014. p. 2801-2808 6900571 (Proceedings of the 2014 IEEE Congress on Evolutionary Computation, CEC 2014).

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

TY - GEN

T1 - Family bootstrapping

T2 - 2014 IEEE Congress on Evolutionary Computation, CEC 2014

AU - Moshaiov, Amiram

AU - Tal, Amir

PY - 2014/9/16

Y1 - 2014/9/16

N2 - Studies on the bootstrap problem in evolutionary robotics help lifting the barrier from the way to evolve robots for complex tasks. It remains an open question, though, how to reduce the need for designer knowledge when devising a bootstrapping approach for any particular complex task. Transfer learning may help reducing this need and support the evolution of solutions to complex tasks, through task relatedness. Relying on the commonalities of similar tasks, we introduce a new concept of Family Bootstrapping (FB). FB refers to the creation of biased ancestors that are expected to onset the evolution of 'a family' of solutions not just for one task, but for a set of related robot tasks. A general FB paradigm is outlined and the unique potential of the proposed concept is discussed. To highlight the validity of the FB concept, a simple demonstration case, concerning the evolution of neuro-controllers for a set of robot navigation tasks, is provided. The paper is concluded with some suggestions for future research.

AB - Studies on the bootstrap problem in evolutionary robotics help lifting the barrier from the way to evolve robots for complex tasks. It remains an open question, though, how to reduce the need for designer knowledge when devising a bootstrapping approach for any particular complex task. Transfer learning may help reducing this need and support the evolution of solutions to complex tasks, through task relatedness. Relying on the commonalities of similar tasks, we introduce a new concept of Family Bootstrapping (FB). FB refers to the creation of biased ancestors that are expected to onset the evolution of 'a family' of solutions not just for one task, but for a set of related robot tasks. A general FB paradigm is outlined and the unique potential of the proposed concept is discussed. To highlight the validity of the FB concept, a simple demonstration case, concerning the evolution of neuro-controllers for a set of robot navigation tasks, is provided. The paper is concluded with some suggestions for future research.

UR - https://www.scopus.com/pages/publications/84908592942

U2 - 10.1109/CEC.2014.6900571

DO - 10.1109/CEC.2014.6900571

M3 - ???researchoutput.researchoutputtypes.contributiontobookanthology.conference???

AN - SCOPUS:84908592942

T3 - Proceedings of the 2014 IEEE Congress on Evolutionary Computation, CEC 2014

SP - 2801

EP - 2808

BT - Proceedings of the 2014 IEEE Congress on Evolutionary Computation, CEC 2014

PB - Institute of Electrical and Electronics Engineers Inc.

Y2 - 6 July 2014 through 11 July 2014

ER -

Moshaiov A, Tal A. Family bootstrapping: A genetic transfer learning approach for onsetting the evolution for a set of related robotic tasks. In Proceedings of the 2014 IEEE Congress on Evolutionary Computation, CEC 2014. Institute of Electrical and Electronics Engineers Inc. 2014. p. 2801-2808. 6900571. (Proceedings of the 2014 IEEE Congress on Evolutionary Computation, CEC 2014). doi: 10.1109/CEC.2014.6900571

Family bootstrapping: A genetic transfer learning approach for onsetting the evolution for a set of related robotic tasks

Abstract

Publication series

Conference

Bibliographical note

Access to Document

Other files and links

Fingerprint

Cite this