Automatic classification and permittivity estimation of organic solvents using a dielectric resonator sensor and machine learning techniques

Abstract

This paper presents the application of a dielectric resonator sensor to characterize organic solvents. Two different acquisition systems were implemented to test the sensor and compare the results between a Vector Network Analyzer (VNA) and a low-cost portable electronic reader presented in this paper. Five dissolutions and air were measured within a permittivity range from 1 to 80. Principal Component Analysis (PCA) and Support Vector Machine (SVM) were used to perform automatic classification achieving an accuracy close to the 100 for both systems.