Constrained optimization of neural network architecture

Mung Chiang

Constrained optimization of neural network architecture

Mung Chiang

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

Abstract

By introducing a well-motivated information theoretic metric and new convex optimization algorithms, the architecture of a neural network is designed to enhance its supervised learning capability. We formulate two optimization frameworks that allow efficient algorithms for a large number of variables and accommodate a variety of practical constraints on structural randomness of neural networks. Convex optimization is also used for independent component analysis (ICA) and multi-antenna fading channel communication channels.

Original language	English (US)
Title of host publication	2000 IEEE Asia-Pacific Conference on Circuits and Systems
Subtitle of host publication	Electronic Communication Systems
Pages	356-359
Number of pages	4
State	Published - Dec 1 2000
Externally published	Yes
Event	2000 IEEE Asia-Pacific Conference on Circuits and Systems: Electronic Communication Systems - Tianjin, China Duration: Dec 4 2000 → Dec 6 2000

Other

Other	2000 IEEE Asia-Pacific Conference on Circuits and Systems: Electronic Communication Systems
Country/Territory	China
City	Tianjin
Period	12/4/00 → 12/6/00

All Science Journal Classification (ASJC) codes

Electrical and Electronic Engineering

Cite this

@inproceedings{3a8b76bb874449279678f85ab51d0a9d,

title = "Constrained optimization of neural network architecture",

abstract = "By introducing a well-motivated information theoretic metric and new convex optimization algorithms, the architecture of a neural network is designed to enhance its supervised learning capability. We formulate two optimization frameworks that allow efficient algorithms for a large number of variables and accommodate a variety of practical constraints on structural randomness of neural networks. Convex optimization is also used for independent component analysis (ICA) and multi-antenna fading channel communication channels.",

author = "Mung Chiang",

year = "2000",

month = dec,

day = "1",

language = "English (US)",

pages = "356--359",

booktitle = "2000 IEEE Asia-Pacific Conference on Circuits and Systems",

note = "2000 IEEE Asia-Pacific Conference on Circuits and Systems: Electronic Communication Systems ; Conference date: 04-12-2000 Through 06-12-2000",

}

TY - GEN

T1 - Constrained optimization of neural network architecture

AU - Chiang, Mung

PY - 2000/12/1

Y1 - 2000/12/1

N2 - By introducing a well-motivated information theoretic metric and new convex optimization algorithms, the architecture of a neural network is designed to enhance its supervised learning capability. We formulate two optimization frameworks that allow efficient algorithms for a large number of variables and accommodate a variety of practical constraints on structural randomness of neural networks. Convex optimization is also used for independent component analysis (ICA) and multi-antenna fading channel communication channels.

AB - By introducing a well-motivated information theoretic metric and new convex optimization algorithms, the architecture of a neural network is designed to enhance its supervised learning capability. We formulate two optimization frameworks that allow efficient algorithms for a large number of variables and accommodate a variety of practical constraints on structural randomness of neural networks. Convex optimization is also used for independent component analysis (ICA) and multi-antenna fading channel communication channels.

UR - http://www.scopus.com/inward/record.url?scp=2342483722&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=2342483722&partnerID=8YFLogxK

M3 - Conference contribution

AN - SCOPUS:2342483722

SP - 356

EP - 359

BT - 2000 IEEE Asia-Pacific Conference on Circuits and Systems

T2 - 2000 IEEE Asia-Pacific Conference on Circuits and Systems: Electronic Communication Systems

Y2 - 4 December 2000 through 6 December 2000

ER -

Constrained optimization of neural network architecture

Abstract

Other

All Science Journal Classification (ASJC) codes

Other files and links

Fingerprint

Cite this