Home Author Index Search Volume 1 May 2009 ISSN 1797-9617

International Journal of

Recent Trends in Engineering

Home > Vol. 1, No. 3


International Journal of Recent Trends in Engineering (IJRTE)

ISSN 1797-9617

Volume 1, Number 3, May 2009

Issue on Electrical & Electronics

Page(s): 140-144

On Comparison of Various Pattern Recognition Techniques for Identifying Escherichia Coli in Clinical Specimens

Subadra M, Marimuthu N. S.

Full text: PDF


The use of Electronic nose (E-nose) technology for detection of bacteria such as Escherichia Coli (E.Coli) has several practical advantages over current laboratory procedures, such as lower cost and reduced testing time. An E-nose consists of a sensory part and a pattern recognition classifier part. The challenging issues in the design of a pattern classifier is to find the optimal architecture and training procedure since it depends on four major issues such as sensitivity, specificity, repeatability and discrimination. In this paper, the sensory data from αFOX 3,000 E-nose at CFTRI, Mysore, for bacterial classification and identification of E.Coli from other samples are taken as data set. This dataset consists of 75 data, out of which 53 subjected for training, 10 for cross validation and 12 for testing. Three different pattern recognition procedures are employed: Multi Layer Perceptron (MLP), Principle Component Analysis hybrid (PCANN), and Support Vector Machine (SVM). Genetic Algorithm is utilized in all these three neural models for optimization. Results indicate that separation between groups can be achieved. Classification accuracy is compared for all procedures and MLP found to exhibit the best performance with the architecture structure as 12-8-3 with Levenberg-Marquardt learning procedure. The overall classification rate, with specificity (correct normal) of 100% and sensitivity (correct abnormal) of 100% was obtained.

Index Terms

Electronic nose (E-Nose), Artificial Neural Network (ANN), Escherichia Coli (E.Coli), Multi Layer Perceptron with Levenberg-Marquardt learning, Genetic Algorithm

Published by Academy Publisher in cooperation with the ACEEE

@ Copyright 2009 ACADEMY PUBLISHER All rights reserved