Date of Award

3-13-1998

Document Type

Thesis

Degree Name

Master of Science

Department

Department of Operational Sciences

First Advisor

Kenneth W. Bauer, PhD

Abstract

The selection of salient features and an appropriate hidden layer architecture contributes significantly to the performance of a neural network. A number of metrics and methodologies exist for estimating these parameters. This research builds on recent efforts to integrate feature and architecture selection for the multilayer perceptron. In the first stage of work a current algorithm is developed in a parallel environment, significantly improving its efficiency and utility. In the second stage, improvements to the algorithm are proposed. With regards to feature selection, a common random number (CRN) addition is proposed. Two new methods of architecture selection are examined, to include an information criterion and a signal to noise based procedure. These methodologies are shown to improve algorithm performance.

AFIT Designator

AFIT-GOR-ENS-98M-20

DTIC Accession Number

ADA342706

Share

COinS