Application of full factorial experiment in designing an ANN-based control chart pattern recognizer

Masood, Ibrahim and Hassan, Adnan (2008) Application of full factorial experiment in designing an ANN-based control chart pattern recognizer. In: International Graduate on Engineering and Science (IGCES'08), 23-24 December, 2008, Universiti Teknologi Malaysia, Sekudai, Johor.

[img]
Preview
PDF
747Kb

Abstract

Automated recognition of control chart patterns for monitoring and diagnosing process quality has been an active area of research since the last 20 years. An artificial neural network (ANN) based models with back-propagation algorithm was known to have resulted the promising recognition accuracy. However, the performance of an ANN depends on a proper selection of the design parameters. In this paper, full factorial design of experiment (DOE) was utilized in investigating several parameters that influence the recognition accuracy of an ANN. This systematic method provided an optimal ANN design with satisfied recognition accuracy.

Item Type:Conference or Workshop Item (Paper)
Subjects:T Technology > T Technology (General)
Divisions:Faculty of Mechanical and Manufacturing Engineering > Department of Engineering Mechanics
ID Code:2221
Deposited By:Normajihan Abd. Rahman
Deposited On:28 Mar 2012 15:57
Last Modified:28 Mar 2012 15:57

Repository Staff Only: item control page