Classification and Gas Concentration Measurements of Human Axillary Odor using Electronic Nose

Human axillary odor produces gas from sweat which concentration will change depends on the activities and metabolism in the body. Sweat concentration can be used as information to determine body health. Nowadays, e-nose is widely used in medicine, food industry, agriculture, and biotechnology. An electronic nose (e-nose) is a device that mimics how the human nose works. This paper will build an e-nose with seven sensors from Figaro Taguchi series (TGS) sensors and one sensor from humidity and temperature sensors (SHT-15 series). The e-nose was used to obtain the human axillary odor in the morning, afternoon, and evening. Several classifiers are used in the classification process and the result showed that Random Forest with tuned hyperparameter produced the best result with an accuracy of 87.43%. By using the ANOVA f-test, it is showed that methane and ethanol from sensor TGS 2612 are the most significant gas in the classification process. The experimental result showed that human axillary odor produced different ethanol and methane gas concentration in the morning, afternoon, and evening.