Abstract
Recent advances in the field of electronic noses (e-noses) have led to new developments in both sensors and feature extraction as well as data processing techniques, providing an increased amount of information. Therefore, feature selection has become essential in the development of e-nose applications. Sophisticated computation techniques can be applied for solving the old problem of sensor number optimization and feature selections. In this way, one can find an optimal application-specific sensor array and reduce the potential cost associated with designing new e-nose devices. In this paper, we examine a procedure to extract and select modeling features for optimal e-nose performance. The usefulness of this approach is demonstrated in detail. We calculated the model’s performance using cross-validation with the standard leave-one-group-out and group shuffle validation methods. Our analysis of wine spoilage data from the sensor array shows when a transient sensor response is considered, both from gas adsorption and desorption phases, it is possible to obtain a reasonable level of odor detection even with data coming from a single sensor. This requires adequate extraction of modeling features and then selection of features used in the final model.
Highlights
Detection and analysis of smells among specified applications can be assessed by many analytical techniques
The four studied wine categories are marked by colors: low-quality wines (LQ)–low quality, average-quality wines (AQ)—average quality, high-quality wines (HQ)—high quality, and Ea—diluted ethanol
The time to train the model required for wrapper methods, which usually provides a better selection of features in terms of model performance, is acceptable on modern computer hardware.In our work, the Mutual Information, Fisher Score, and RelfiefF methods [49] were used for comparisons with recursive forward selection method described above
Summary
Detection and analysis of smells among specified applications can be assessed by many analytical techniques. Promising is surface plasmon resonance imaging [2] and its successful application for gas-phase detection of volatile organic compounds [3] These new electronic instrumentations are capable of imitating the remarkable abilities of the human nose, and they have proved their feasibility and effectiveness in odor recognition, environmental monitoring [4], medical diagnosis [5], as well as food quality monitoring [6,7,8,9]. Evolutionary computation techniques can be applied to optimize sensor selection, feature selection, and classification stages [23] In this way, one can find an optimal subset of sensors for a particular application while choosing sensing devices from a larger database of sensors.
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Free) 
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 call
