Demonstration of Intelligent Sensing by Nanosensors and Use of Classification and Regression Models for Electronic Nose Applications

Electronic nose (E-Nose) comprises a sensor array combined with scientific computations to derive intelligent analysis. This work presents an E-Nose consisting of four nanosensors and performs smart analysis with machine learning (ML) models for gas type classification and concentration prediction. The sensors were tested with ethanol, acetone, isopropyl alcohol (IPA), xylene, benzene, and toluene gases, and results showed higher selectivity to ethanol and acetone. The combination of principal component analysis (PCA) and support vector classifier (SVC) having radial basis function (RBF) as kernel classified ethanol, acetone, and other tested gases with 98% accuracy. Furthermore, the regression studies performed with AdaBoost algorithm yielded <inline-formula xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink"> <tex-math notation="LaTeX">${R}2$ </tex-math></inline-formula> scores of 80% and 83% for ethanol and acetone concentration prediction, respectively.