Advanced statistical metrics for gas identification system with quantification feedback

Samir Brahim-Belhaouari, Muhammad Hassan, Nicolas Walter, Amine Bermak

Research output: Contribution to journalArticle

10 Citations (Scopus)

Abstract

The pattern recognition problem for real-life applications of gas identification is challenging due to the limited amount of data existing and the sequential variability of the mechanism mostly caused by drift and the real-time detection. These problems are commonly caused by the slow response of most of the gas sensors. In this paper, a novel gas identification approach based on the cluster-k-nearest neighbor (C-k-NN) is introduced. The effectiveness of this approach has been successfully demonstrated on the experimental data set obtained from array of gas sensors. Our classification takes advantages of both the k-NN, which is highly accurate, and the k-means cluster, which is able to reduce the classification time. In order to increase the accuracy rate, a new feature selection method is proposed. The selection of features is based on their ability to separate and distinguish between different classes. Advanced statistical metrics are introduced to quantify the classification contribution of each feature. Mostly, classifiers are suffering from misclassification detection; new statistical metrics are introduced to estimate the exactness of the classifier response, i.e., to detect the misclassification. To enhance the classification performances for gas identification, a new tree classification design is introduced, named tree C-k-NN. In order to assess the technique, experiments were conducted on six different gases. Accuracy rate of 98.7% has been obtained with the C-k-NN and 100% with the tree C-k-NN. The performance of this approach is also validated using three publicly available data sets.

Original languageEnglish
Article number6934986
Pages (from-to)1705-1715
Number of pages11
JournalIEEE Sensors Journal
Volume15
Issue number3
DOIs
Publication statusPublished - 1 Mar 2015
Externally publishedYes

Fingerprint

system identification
Identification (control systems)
Feedback
Gases
gases
Chemical sensors
classifiers
Classifiers
sensors
Pattern recognition
Feature extraction
pattern recognition
estimates
Experiments

Keywords

  • classification
  • Clustering
  • Confidence coefficient
  • feature selection
  • gas identification
  • Gaussian Mixture Model (GMM)
  • K-Means

ASJC Scopus subject areas

  • Instrumentation
  • Electrical and Electronic Engineering

Cite this

Advanced statistical metrics for gas identification system with quantification feedback. / Brahim-Belhaouari, Samir; Hassan, Muhammad; Walter, Nicolas; Bermak, Amine.

In: IEEE Sensors Journal, Vol. 15, No. 3, 6934986, 01.03.2015, p. 1705-1715.

Research output: Contribution to journalArticle

Brahim-Belhaouari, Samir ; Hassan, Muhammad ; Walter, Nicolas ; Bermak, Amine. / Advanced statistical metrics for gas identification system with quantification feedback. In: IEEE Sensors Journal. 2015 ; Vol. 15, No. 3. pp. 1705-1715.
@article{6ef620eacd0948aeb5419fbeb51d68bd,
title = "Advanced statistical metrics for gas identification system with quantification feedback",
abstract = "The pattern recognition problem for real-life applications of gas identification is challenging due to the limited amount of data existing and the sequential variability of the mechanism mostly caused by drift and the real-time detection. These problems are commonly caused by the slow response of most of the gas sensors. In this paper, a novel gas identification approach based on the cluster-k-nearest neighbor (C-k-NN) is introduced. The effectiveness of this approach has been successfully demonstrated on the experimental data set obtained from array of gas sensors. Our classification takes advantages of both the k-NN, which is highly accurate, and the k-means cluster, which is able to reduce the classification time. In order to increase the accuracy rate, a new feature selection method is proposed. The selection of features is based on their ability to separate and distinguish between different classes. Advanced statistical metrics are introduced to quantify the classification contribution of each feature. Mostly, classifiers are suffering from misclassification detection; new statistical metrics are introduced to estimate the exactness of the classifier response, i.e., to detect the misclassification. To enhance the classification performances for gas identification, a new tree classification design is introduced, named tree C-k-NN. In order to assess the technique, experiments were conducted on six different gases. Accuracy rate of 98.7{\%} has been obtained with the C-k-NN and 100{\%} with the tree C-k-NN. The performance of this approach is also validated using three publicly available data sets.",
keywords = "classification, Clustering, Confidence coefficient, feature selection, gas identification, Gaussian Mixture Model (GMM), K-Means",
author = "Samir Brahim-Belhaouari and Muhammad Hassan and Nicolas Walter and Amine Bermak",
year = "2015",
month = "3",
day = "1",
doi = "10.1109/JSEN.2014.2364687",
language = "English",
volume = "15",
pages = "1705--1715",
journal = "IEEE Sensors Journal",
issn = "1530-437X",
publisher = "Institute of Electrical and Electronics Engineers Inc.",
number = "3",

}

TY - JOUR

T1 - Advanced statistical metrics for gas identification system with quantification feedback

AU - Brahim-Belhaouari, Samir

AU - Hassan, Muhammad

AU - Walter, Nicolas

AU - Bermak, Amine

PY - 2015/3/1

Y1 - 2015/3/1

N2 - The pattern recognition problem for real-life applications of gas identification is challenging due to the limited amount of data existing and the sequential variability of the mechanism mostly caused by drift and the real-time detection. These problems are commonly caused by the slow response of most of the gas sensors. In this paper, a novel gas identification approach based on the cluster-k-nearest neighbor (C-k-NN) is introduced. The effectiveness of this approach has been successfully demonstrated on the experimental data set obtained from array of gas sensors. Our classification takes advantages of both the k-NN, which is highly accurate, and the k-means cluster, which is able to reduce the classification time. In order to increase the accuracy rate, a new feature selection method is proposed. The selection of features is based on their ability to separate and distinguish between different classes. Advanced statistical metrics are introduced to quantify the classification contribution of each feature. Mostly, classifiers are suffering from misclassification detection; new statistical metrics are introduced to estimate the exactness of the classifier response, i.e., to detect the misclassification. To enhance the classification performances for gas identification, a new tree classification design is introduced, named tree C-k-NN. In order to assess the technique, experiments were conducted on six different gases. Accuracy rate of 98.7% has been obtained with the C-k-NN and 100% with the tree C-k-NN. The performance of this approach is also validated using three publicly available data sets.

AB - The pattern recognition problem for real-life applications of gas identification is challenging due to the limited amount of data existing and the sequential variability of the mechanism mostly caused by drift and the real-time detection. These problems are commonly caused by the slow response of most of the gas sensors. In this paper, a novel gas identification approach based on the cluster-k-nearest neighbor (C-k-NN) is introduced. The effectiveness of this approach has been successfully demonstrated on the experimental data set obtained from array of gas sensors. Our classification takes advantages of both the k-NN, which is highly accurate, and the k-means cluster, which is able to reduce the classification time. In order to increase the accuracy rate, a new feature selection method is proposed. The selection of features is based on their ability to separate and distinguish between different classes. Advanced statistical metrics are introduced to quantify the classification contribution of each feature. Mostly, classifiers are suffering from misclassification detection; new statistical metrics are introduced to estimate the exactness of the classifier response, i.e., to detect the misclassification. To enhance the classification performances for gas identification, a new tree classification design is introduced, named tree C-k-NN. In order to assess the technique, experiments were conducted on six different gases. Accuracy rate of 98.7% has been obtained with the C-k-NN and 100% with the tree C-k-NN. The performance of this approach is also validated using three publicly available data sets.

KW - classification

KW - Clustering

KW - Confidence coefficient

KW - feature selection

KW - gas identification

KW - Gaussian Mixture Model (GMM)

KW - K-Means

UR - http://www.scopus.com/inward/record.url?scp=84921059027&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84921059027&partnerID=8YFLogxK

U2 - 10.1109/JSEN.2014.2364687

DO - 10.1109/JSEN.2014.2364687

M3 - Article

AN - SCOPUS:84921059027

VL - 15

SP - 1705

EP - 1715

JO - IEEE Sensors Journal

JF - IEEE Sensors Journal

SN - 1530-437X

IS - 3

M1 - 6934986

ER -