Key points are not available for this paper at this time.
Due to numerous breakthroughs in the domains of gas sensors and machine learning (ML) technology, electronic nose (EN) recently became an important instrument for the invasive-free diagnosis of human and plant diseases. ML algorithms detect disease-causing pathogens in plants and forecast human ailments using the EN data as input. The review of the literature that has already been published on ML techniques employed in EN for the healthcare, agriculture, and other allied domains is presented in this study. Using various search parameters, we conducted a thorough literature search in the online IEEE Xplore, PubMed, ScienceDirect, Springer Link, Google Scholar, and Web of Science databases for a time span of 15 years from 2007 to 2022 pertaining to the subject at hand. While the majority of research has employed commercially accessible EN devices and metal-oxide sensor (MOS), the most common classification techniques include support vector machine (SVM), random forest (RF), and artificial neural network (ANN), along with principal component analysis (PCA) or linear discriminant analysis (LDA) as a feature transformation method. Therefore, this work will serve as a stepping stone for further research in the area of developing novel EN devices and ML algorithms for improving prediction tasks in the domains of healthcare, agriculture, and other allied domains.
Baruah et al. (Wed,) studied this question.