Abstract
For electronic nose systems to obtain meaningful information from sensor data, sensor response features are first extracted for further signal processing. However, redundant features may diminish the accuracy of gas classification. To solve this problem, a minimum distance inlier probability (MDIP) feature selection (FS) method is proposed. By incorporating the intrinsic properties of features and ranking strategy, MDIP can efficiently eliminate redundant features and provide better classification accuracy. The performance of the method was validated on two open-access datasets that provide information for system variation and sensor drift problems, respectively. Experimental results revealed that the average classification accuracy for the two datasets was higher by 46.1% and 37.5%, respectively, with the MDIP method.
Highlights
An electronic nose (E-nose) emulates the human olfactory system, which identifies gases, provides perceptions, and makes judgments
The minimum distance inlier probability (MDIP) outperformed linear support vector machine recursive feature elimination (SVM-RFE) in all batches except batches 7 and 10. This probably resulted because batches 7 and 10 had few meaningful features in common with batch 1 selected by the MDIP method
For dataset 1, the MDIP method achieved the highest accuracy at Tsa = 0.4 (Fig. 2); the smallest standard deviation was observed at Tsa = 0.1
Summary
An electronic nose (E-nose) emulates the human olfactory system, which identifies gases, provides perceptions, and makes judgments. Many studies show that incorporating feature selection methods into E-nose’s signal processing procedure can significantly improve gas classification accuracy. Wrappers use a classifier to evaluate subsets according to their predictive accuracy through statistical resampling or cross validation [13], [14], [16]. Embedded methods, such as support vector machine recursive feature elimination (SVM-RFE), combine filters and wrappers [17]. Tang: MDIP FS Method to Improve Gas Classification for E-Nose Systems. The MDIP method is not limited to a specific classifier, and it assesses features by considering two feature properties: the separation and the reproducibility. Propose a ranking strategy that evaluates the features by considering reproducibility and separability simultaneously to find the most representative features
Sign-in/Register to view highlights & summary Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a callDisclaimer: All third-party content on this website/platform is and will remain the property of their respective owners and is provided on "as is" basis without any warranties, express or implied. Use of third-party content does not indicate any affiliation, sponsorship with or endorsement by them. Any references to third-party content is to identify the corresponding services and shall be considered fair use under The CopyrightLaw.
