Combining an Electronic Nose System with Artificial Intelligence Methods for Odor Identification of Spirits

This study uses the Electronic Nose (E-nose) System we developed to characterize the odor profiles of Volatile Organic Compounds (VOCs) in different types of spirits samples. The performance results of different Machine Learning models are also tested. In this study, the performance of AI classifier models was evaluated, and the Convolutional Neural Network (CNN) model combined with Long Short-Term Memory (LSTM) achieved the best performance, with a loss of 0.03 and an accuracy of 0.98. Preliminary results from this study demonstrate that the electronic nose system we developed can achieve excellent results.