Support Vector Regression on Near-Infrared Spectra Analysis for Moisture Determination in Lignitic Coal Samples

Abstract

We studied rapid moisture determination in lignitic coal samples using near-infrared (NIR) spectrometry technique. This research applied support vector regression (SVR) and discrete wavelet transform (DWT) to analyze NIR spectra. Firstly, NIR spectra were pre-processed by DWT for fitting and compression. Then, DWT coefficients were used to build support vector regression model. Through parameters optimization, the results show that DWT-SVR can obtain satisfactory performance for moisture determination in lignitic coal samples.