Comparative Analysis of KNN, Random Forrest, Decision Tree and SVM in Application to Millimeter Wave Chemical Detection

With the goal of mitigating the potential risks of the undetected hazardous chemicals in public or private places this study proposes to establish a system of detecting hazardous chemicals in standoff environment using millimeter wave and machine learning. Higher accuracy and speed in detection of hazardous chemicals though increased sensitivity and selectivity must be improved. With the minimal or no false alarm, this will boost the reliability of the system. The aim of this study is to develop a system that will provide real-time detection of hazardous chemicals at a distance of at least 1 meter. The use of millimeter wave at a frequency of 77-81MHz will identify the spectral content of the chemical. Identification is enhanced with the use of the best model in machine learning tool.