Identification of liquor adulteration based on machine learning and electrochemical sensor

This study introduces a novel approach to detecting liquor adulteration using machine learning algorithms in conjunction with electrochemical sensors. The multi-frequency large-amplitude pulse voltammetric method was used to evaluate six types of liquors (L-50, L-60, L-70, L-80, L-90, L-100) using different metal electrodes at varying frequency segments. It was found that platinum and gold electrodes could distinguish the six types of liquors at the 10 Hz and 100 Hz frequency segments, respectively. An Extreme Learning Machine (ELM) model was established, achieving a classification accuracy rate of 90.0% when the number of hidden layer nodes was 36. Finally, the predictive performance of the ABC-LSSVM model was compared with GSA-LSSVM, GA-LSSVM, and PSO-LSSVM models, with the ABC-LSSVM model demonstrating superior predictive ability for liquor purity.