An Empirical Study for PCA- and LDA-Based Feature Reduction for Gas Identification

Muhammad Ali Akbar, Amine Ait Si Ali, Abbes Amira, Faycal Bensaali, Mohieddine Benammar, Muhammad Hassan, Amine Bermak

Research output: Contribution to journalArticle

23 Citations (Scopus)

Abstract

Increasing the number of sensors in a gas identification system generally improves its performance as this will add extra features for analysis. However, this affects the computational complexity, especially if the identification algorithm is to be implemented on a hardware platform. Therefore, feature reduction is required to extract the most important information from the sensors for processing. In this paper, linear discriminant analysis (LDA) and principal component analysis (PCA)-based feature reduction algorithms have been analyzed using the data obtained from two different types of gas sensors, i.e., seven commercial Figaro sensors and in-house fabricated $4 \times 4$ tin-oxide gas array sensor. A decision tree-based classifier is used to examine the performance of both the PCA and LDA approaches. The software implementation is carried out in MATLAB and the hardware implementation is performed using the Zynq system-on-chip (SoC) platform. It has been found that with the $4 \times 4$ array sensor, two discriminant functions (DF) of LDA provide 3.3% better classification than five PCA components, while for the seven Figaro sensors, two principal components and one DF show the same performances. The hardware implementation results on the programmable logic of the Zynq SoC shows that LDA outperforms PCA by using 50% less resources as well as by being 11% faster with a maximum running frequency of 122 MHz.

Original languageEnglish
Article number7467450
Pages (from-to)5734-5746
Number of pages13
JournalIEEE Sensors Journal
Volume16
Issue number14
DOIs
Publication statusPublished - 15 Jul 2016

Fingerprint

Discriminant analysis
principal components analysis
Principal component analysis
Identification (control systems)
sensors
Sensors
Sensor arrays
Gases
Hardware
gases
hardware
Decision trees
Tin oxides
Chemical sensors
MATLAB
platforms
chips
Computational complexity
Classifiers
system identification

Keywords

  • Electronic nose
  • Feature reduction
  • Gas identification
  • LDA
  • PCA
  • Zynq SoC

ASJC Scopus subject areas

  • Electrical and Electronic Engineering
  • Instrumentation

Cite this

Akbar, M. A., Ait Si Ali, A., Amira, A., Bensaali, F., Benammar, M., Hassan, M., & Bermak, A. (2016). An Empirical Study for PCA- and LDA-Based Feature Reduction for Gas Identification. IEEE Sensors Journal, 16(14), 5734-5746. [7467450]. https://doi.org/10.1109/JSEN.2016.2565721

An Empirical Study for PCA- and LDA-Based Feature Reduction for Gas Identification. / Akbar, Muhammad Ali; Ait Si Ali, Amine; Amira, Abbes; Bensaali, Faycal; Benammar, Mohieddine; Hassan, Muhammad; Bermak, Amine.

In: IEEE Sensors Journal, Vol. 16, No. 14, 7467450, 15.07.2016, p. 5734-5746.

Research output: Contribution to journalArticle

Akbar, MA, Ait Si Ali, A, Amira, A, Bensaali, F, Benammar, M, Hassan, M & Bermak, A 2016, 'An Empirical Study for PCA- and LDA-Based Feature Reduction for Gas Identification', IEEE Sensors Journal, vol. 16, no. 14, 7467450, pp. 5734-5746. https://doi.org/10.1109/JSEN.2016.2565721
Akbar MA, Ait Si Ali A, Amira A, Bensaali F, Benammar M, Hassan M et al. An Empirical Study for PCA- and LDA-Based Feature Reduction for Gas Identification. IEEE Sensors Journal. 2016 Jul 15;16(14):5734-5746. 7467450. https://doi.org/10.1109/JSEN.2016.2565721
Akbar, Muhammad Ali ; Ait Si Ali, Amine ; Amira, Abbes ; Bensaali, Faycal ; Benammar, Mohieddine ; Hassan, Muhammad ; Bermak, Amine. / An Empirical Study for PCA- and LDA-Based Feature Reduction for Gas Identification. In: IEEE Sensors Journal. 2016 ; Vol. 16, No. 14. pp. 5734-5746.
@article{55b780a6f8a542708559069169365f00,
title = "An Empirical Study for PCA- and LDA-Based Feature Reduction for Gas Identification",
abstract = "Increasing the number of sensors in a gas identification system generally improves its performance as this will add extra features for analysis. However, this affects the computational complexity, especially if the identification algorithm is to be implemented on a hardware platform. Therefore, feature reduction is required to extract the most important information from the sensors for processing. In this paper, linear discriminant analysis (LDA) and principal component analysis (PCA)-based feature reduction algorithms have been analyzed using the data obtained from two different types of gas sensors, i.e., seven commercial Figaro sensors and in-house fabricated $4 \times 4$ tin-oxide gas array sensor. A decision tree-based classifier is used to examine the performance of both the PCA and LDA approaches. The software implementation is carried out in MATLAB and the hardware implementation is performed using the Zynq system-on-chip (SoC) platform. It has been found that with the $4 \times 4$ array sensor, two discriminant functions (DF) of LDA provide 3.3{\%} better classification than five PCA components, while for the seven Figaro sensors, two principal components and one DF show the same performances. The hardware implementation results on the programmable logic of the Zynq SoC shows that LDA outperforms PCA by using 50{\%} less resources as well as by being 11{\%} faster with a maximum running frequency of 122 MHz.",
keywords = "Electronic nose, Feature reduction, Gas identification, LDA, PCA, Zynq SoC",
author = "Akbar, {Muhammad Ali} and {Ait Si Ali}, Amine and Abbes Amira and Faycal Bensaali and Mohieddine Benammar and Muhammad Hassan and Amine Bermak",
year = "2016",
month = "7",
day = "15",
doi = "10.1109/JSEN.2016.2565721",
language = "English",
volume = "16",
pages = "5734--5746",
journal = "IEEE Sensors Journal",
issn = "1530-437X",
publisher = "Institute of Electrical and Electronics Engineers Inc.",
number = "14",

}

TY - JOUR

T1 - An Empirical Study for PCA- and LDA-Based Feature Reduction for Gas Identification

AU - Akbar, Muhammad Ali

AU - Ait Si Ali, Amine

AU - Amira, Abbes

AU - Bensaali, Faycal

AU - Benammar, Mohieddine

AU - Hassan, Muhammad

AU - Bermak, Amine

PY - 2016/7/15

Y1 - 2016/7/15

N2 - Increasing the number of sensors in a gas identification system generally improves its performance as this will add extra features for analysis. However, this affects the computational complexity, especially if the identification algorithm is to be implemented on a hardware platform. Therefore, feature reduction is required to extract the most important information from the sensors for processing. In this paper, linear discriminant analysis (LDA) and principal component analysis (PCA)-based feature reduction algorithms have been analyzed using the data obtained from two different types of gas sensors, i.e., seven commercial Figaro sensors and in-house fabricated $4 \times 4$ tin-oxide gas array sensor. A decision tree-based classifier is used to examine the performance of both the PCA and LDA approaches. The software implementation is carried out in MATLAB and the hardware implementation is performed using the Zynq system-on-chip (SoC) platform. It has been found that with the $4 \times 4$ array sensor, two discriminant functions (DF) of LDA provide 3.3% better classification than five PCA components, while for the seven Figaro sensors, two principal components and one DF show the same performances. The hardware implementation results on the programmable logic of the Zynq SoC shows that LDA outperforms PCA by using 50% less resources as well as by being 11% faster with a maximum running frequency of 122 MHz.

AB - Increasing the number of sensors in a gas identification system generally improves its performance as this will add extra features for analysis. However, this affects the computational complexity, especially if the identification algorithm is to be implemented on a hardware platform. Therefore, feature reduction is required to extract the most important information from the sensors for processing. In this paper, linear discriminant analysis (LDA) and principal component analysis (PCA)-based feature reduction algorithms have been analyzed using the data obtained from two different types of gas sensors, i.e., seven commercial Figaro sensors and in-house fabricated $4 \times 4$ tin-oxide gas array sensor. A decision tree-based classifier is used to examine the performance of both the PCA and LDA approaches. The software implementation is carried out in MATLAB and the hardware implementation is performed using the Zynq system-on-chip (SoC) platform. It has been found that with the $4 \times 4$ array sensor, two discriminant functions (DF) of LDA provide 3.3% better classification than five PCA components, while for the seven Figaro sensors, two principal components and one DF show the same performances. The hardware implementation results on the programmable logic of the Zynq SoC shows that LDA outperforms PCA by using 50% less resources as well as by being 11% faster with a maximum running frequency of 122 MHz.

KW - Electronic nose

KW - Feature reduction

KW - Gas identification

KW - LDA

KW - PCA

KW - Zynq SoC

UR - http://www.scopus.com/inward/record.url?scp=84976468221&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84976468221&partnerID=8YFLogxK

U2 - 10.1109/JSEN.2016.2565721

DO - 10.1109/JSEN.2016.2565721

M3 - Article

AN - SCOPUS:84976468221

VL - 16

SP - 5734

EP - 5746

JO - IEEE Sensors Journal

JF - IEEE Sensors Journal

SN - 1530-437X

IS - 14

M1 - 7467450

ER -