The MI - RBFN: Mapping for generalization

Paul B. Deignan; Peter H. Meckl; Matthew A. Franchek

doi:10.1109/acc.2002.1024527

The MI - RBFN: Mapping for generalization

Paul B. Deignan, Peter H. Meckl, Matthew A. Franchek

Research Institute

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

3 Scopus citations

Abstract

The mutual information - radial basis function network (MI-RBFN) is an efficient, general, and integrated method of approximating complex, continuous, deterministic systems from incomplete information. The nodes of the MI-RBFN are located by clustering local mutual information estimates thereby yielding a mapping that inherently generalizes better than one formulated by seeking solely to minimize residuals. The expectation-maximization algorithm is introduced for Gaussian clustering of MI estimates. A further improvement in the methodology is marked by the specification of a set of rules for intelligently determining the binning interval of the input and target spaces.

Original language	English (US)
Pages (from-to)	3840-3845
Number of pages	6
Journal	Proceedings of the American Control Conference
Volume	5
DOIs	https://doi.org/10.1109/acc.2002.1024527
State	Published - 2002
Event	2002 American Control Conference - Anchorage, AK, United States Duration: May 8 2002 → May 10 2002

ASJC Scopus subject areas

Electrical and Electronic Engineering

Access to Document

10.1109/acc.2002.1024527

Cite this

@article{d98262b05769472f83426cb99446187e,

title = "The MI - RBFN: Mapping for generalization",

abstract = "The mutual information - radial basis function network (MI-RBFN) is an efficient, general, and integrated method of approximating complex, continuous, deterministic systems from incomplete information. The nodes of the MI-RBFN are located by clustering local mutual information estimates thereby yielding a mapping that inherently generalizes better than one formulated by seeking solely to minimize residuals. The expectation-maximization algorithm is introduced for Gaussian clustering of MI estimates. A further improvement in the methodology is marked by the specification of a set of rules for intelligently determining the binning interval of the input and target spaces.",

author = "Deignan, {Paul B.} and Meckl, {Peter H.} and Franchek, {Matthew A.}",

year = "2002",

doi = "10.1109/acc.2002.1024527",

language = "English (US)",

volume = "5",

pages = "3840--3845",

journal = "Proceedings of the American Control Conference",

issn = "0743-1619",

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

}

TY - JOUR

T1 - The MI - RBFN

T2 - 2002 American Control Conference

AU - Deignan, Paul B.

AU - Meckl, Peter H.

AU - Franchek, Matthew A.

PY - 2002

Y1 - 2002

N2 - The mutual information - radial basis function network (MI-RBFN) is an efficient, general, and integrated method of approximating complex, continuous, deterministic systems from incomplete information. The nodes of the MI-RBFN are located by clustering local mutual information estimates thereby yielding a mapping that inherently generalizes better than one formulated by seeking solely to minimize residuals. The expectation-maximization algorithm is introduced for Gaussian clustering of MI estimates. A further improvement in the methodology is marked by the specification of a set of rules for intelligently determining the binning interval of the input and target spaces.

AB - The mutual information - radial basis function network (MI-RBFN) is an efficient, general, and integrated method of approximating complex, continuous, deterministic systems from incomplete information. The nodes of the MI-RBFN are located by clustering local mutual information estimates thereby yielding a mapping that inherently generalizes better than one formulated by seeking solely to minimize residuals. The expectation-maximization algorithm is introduced for Gaussian clustering of MI estimates. A further improvement in the methodology is marked by the specification of a set of rules for intelligently determining the binning interval of the input and target spaces.

UR - http://www.scopus.com/inward/record.url?scp=0036055906&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=0036055906&partnerID=8YFLogxK

U2 - 10.1109/acc.2002.1024527

DO - 10.1109/acc.2002.1024527

M3 - Conference article

AN - SCOPUS:0036055906

VL - 5

SP - 3840

EP - 3845

JO - Proceedings of the American Control Conference

JF - Proceedings of the American Control Conference

SN - 0743-1619

Y2 - 8 May 2002 through 10 May 2002

ER -

The MI - RBFN: Mapping for generalization

Abstract

ASJC Scopus subject areas

Access to Document

Other files and links

Fingerprint

Cite this