Stochastic information gradient algorithm based on maximum entropy density estimation

Badong Chen; Yu Zhu; Jinchun Hu; Ming Zhang

Stochastic information gradient algorithm based on maximum entropy density estimation

Badong Chen
, Yu Zhu
, Jinchun Hu
, Ming Zhang

Tsinghua University

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

7 Scopus citations

Abstract

We propose a new stochastic information gradient (SIG) algorithm based on an online maximum entropy density estimation of the error distribution. The proposed algorithm is simple, yet efficient in implementation, as it involves no choice of bandwidth and does not resort to Newton's method of optimization. Simulation results demonstrate the favorable performance of the new algorithm. ICIC International

Original language	English
Pages (from-to)	1141-1145
Number of pages	5
Journal	ICIC Express Letters
Volume	4
Issue number	4
State	Published - Aug 2010
Externally published	Yes

Cite this

@article{0c030c70d4974f948cccffa484fa712c,

title = "Stochastic information gradient algorithm based on maximum entropy density estimation",

abstract = "We propose a new stochastic information gradient (SIG) algorithm based on an online maximum entropy density estimation of the error distribution. The proposed algorithm is simple, yet efficient in implementation, as it involves no choice of bandwidth and does not resort to Newton's method of optimization. Simulation results demonstrate the favorable performance of the new algorithm. ICIC International",

author = "Badong Chen and Yu Zhu and Jinchun Hu and Ming Zhang",

year = "2010",

month = aug,

language = "英语",

volume = "4",

pages = "1141--1145",

journal = "ICIC Express Letters",

issn = "1881-803X",

publisher = "ICIC International",

number = "4",

}

TY - JOUR

T1 - Stochastic information gradient algorithm based on maximum entropy density estimation

AU - Chen, Badong

AU - Zhu, Yu

AU - Hu, Jinchun

AU - Zhang, Ming

PY - 2010/8

Y1 - 2010/8

N2 - We propose a new stochastic information gradient (SIG) algorithm based on an online maximum entropy density estimation of the error distribution. The proposed algorithm is simple, yet efficient in implementation, as it involves no choice of bandwidth and does not resort to Newton's method of optimization. Simulation results demonstrate the favorable performance of the new algorithm. ICIC International

AB - We propose a new stochastic information gradient (SIG) algorithm based on an online maximum entropy density estimation of the error distribution. The proposed algorithm is simple, yet efficient in implementation, as it involves no choice of bandwidth and does not resort to Newton's method of optimization. Simulation results demonstrate the favorable performance of the new algorithm. ICIC International

UR - https://www.scopus.com/pages/publications/77955208429

M3 - 文章

AN - SCOPUS:77955208429

SN - 1881-803X

VL - 4

SP - 1141

EP - 1145

JO - ICIC Express Letters

JF - ICIC Express Letters

IS - 4

ER -

Stochastic information gradient algorithm based on maximum entropy density estimation

Abstract

Other files and links

Fingerprint

Cite this