Stochastic backpropagation: A learning algorithm for generalization problems

C. V. Ramamoorthy; Shashi Shekhar

Stochastic backpropagation: A learning algorithm for generalization problems

C. V. Ramamoorthy, Shashi Shekhar

Computer Science and Engineering

Research output: Contribution to journal › Conference article › peer-review

4 Scopus citations

Abstract

Neural networks have traditionally been applied to recognition problems, and most learning algorithms are tailored to those problems. The authors discuss the requirements of learning for generalization, which is NP-complete and cannot be approached by traditional methods based on gradient descent. They present a stochastic learning algorithm based on simulated annealing in weight space. The convergence properties and feasibility of the algorithm are verified.

Original language	English (US)
Pages (from-to)	664-671
Number of pages	8
Journal	Proceedings - IEEE Computer Society's International Computer Software & Applications Conference
State	Published - Dec 1 1989
Event	Proceedings of the Thirteenth Annual International Computer Software & Applications Conference - COMPSAC 89 - Orlando, FL, USA Duration: Sep 20 1989 → Sep 22 1989

OpenUrl availability

Full text

Cite this

@article{7cea11149ad340828aa9cf27c3a8c5af,

title = "Stochastic backpropagation: A learning algorithm for generalization problems",

abstract = "Neural networks have traditionally been applied to recognition problems, and most learning algorithms are tailored to those problems. The authors discuss the requirements of learning for generalization, which is NP-complete and cannot be approached by traditional methods based on gradient descent. They present a stochastic learning algorithm based on simulated annealing in weight space. The convergence properties and feasibility of the algorithm are verified.",

author = "Ramamoorthy, {C. V.} and Shashi Shekhar",

year = "1989",

month = dec,

day = "1",

language = "English (US)",

pages = "664--671",

journal = "Proceedings - IEEE Computer Society's International Computer Software & Applications Conference",

issn = "0730-6512",

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

note = "Proceedings of the Thirteenth Annual International Computer Software & Applications Conference - COMPSAC 89 ; Conference date: 20-09-1989 Through 22-09-1989",

}

TY - JOUR

T1 - Stochastic backpropagation

T2 - Proceedings of the Thirteenth Annual International Computer Software & Applications Conference - COMPSAC 89

AU - Ramamoorthy, C. V.

AU - Shekhar, Shashi

PY - 1989/12/1

Y1 - 1989/12/1

N2 - Neural networks have traditionally been applied to recognition problems, and most learning algorithms are tailored to those problems. The authors discuss the requirements of learning for generalization, which is NP-complete and cannot be approached by traditional methods based on gradient descent. They present a stochastic learning algorithm based on simulated annealing in weight space. The convergence properties and feasibility of the algorithm are verified.

AB - Neural networks have traditionally been applied to recognition problems, and most learning algorithms are tailored to those problems. The authors discuss the requirements of learning for generalization, which is NP-complete and cannot be approached by traditional methods based on gradient descent. They present a stochastic learning algorithm based on simulated annealing in weight space. The convergence properties and feasibility of the algorithm are verified.

UR - http://www.scopus.com/inward/record.url?scp=0024883923&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=0024883923&partnerID=8YFLogxK

M3 - Conference article

AN - SCOPUS:0024883923

SN - 0730-6512

SP - 664

EP - 671

JO - Proceedings - IEEE Computer Society's International Computer Software & Applications Conference

JF - Proceedings - IEEE Computer Society's International Computer Software & Applications Conference

Y2 - 20 September 1989 through 22 September 1989

ER -

Stochastic backpropagation: A learning algorithm for generalization problems

Abstract

OpenUrl availability

Other files and links

Fingerprint

Cite this