Share Email Print

Proceedings Paper

Software metric-based neural network classification models of a very large telecommunications system
Author(s): Taghi M. Khoshgoftaar; Edward B. Allen; John Hudepohl; Steve Aud
Format Member Price Non-Member Price
PDF $17.00 $21.00

Paper Abstract

Society relies on telecommunications to such an extent that telecommunications software must have high reliability. Neural network models can be used to identify fault-prone modules for extra attention early in development, and thus reduce the risk of unexpected problems with those modules. This paper is an experience report on a model of a large telecommunications system with almost 7,000 changed modules, consisting of over 7 million lines of code in procedures. We developed a neural network model to identify fault-prone modules based on nine design product metrics. Misclassification of not fault-prone modules would incur only modest cost in terms of extra attention to those modules. Misclassification of fault-prone modules would risk unexpected problems late in the development or even after release. Changed modules were randomly divided into a fit data set and a validate data set. We simulated utilization of the fitted model with the validate data set, successfully demonstrating generalization.

Paper Details

Date Published: 22 March 1996
PDF: 12 pages
Proc. SPIE 2760, Applications and Science of Artificial Neural Networks II, (22 March 1996); doi: 10.1117/12.235967
Show Author Affiliations
Taghi M. Khoshgoftaar, Florida Atlantic Univ. (United States)
Edward B. Allen, Florida Atlantic Univ. (United States)
John Hudepohl, Bell Northern Research Inc. (United States)
Steve Aud, Bell Northern Research Inc. (United States)

Published in SPIE Proceedings Vol. 2760:
Applications and Science of Artificial Neural Networks II
Steven K. Rogers; Dennis W. Ruck, Editor(s)

© SPIE. Terms of Use
Back to Top