Feature Wavelength Selection in Near-Infrared Spectroscopy Based on Genetic Algorithm

Feature wavelength selection is similar to dimensionality reduction technique, which aims to reduce the number of variables involved in modeling because they are redundant, uninformative, interfering, or accidentally correlated. Thus, feature wavelength selection has become a significant step for the modeling process of near infrared (NIR) spectral analysis. In this paper, genetic algorithm (GA) for feature wavelength selection is proposed. GA randomly simulates the natural selection of biological evolution to achieve global optimization and selects the optimal feature subset from the variable space. Partial least squares (PLS) algorithm is used for establishing regression analysis prediction model. Comparing with successive projections algorithm (SPA) and competitive adaptive reweighted sampling (CARS), the quantitative model based on GA has better prediction ability and robustness. The correlation coefficient of prediction (RP), the root mean square errors of prediction (RMSEP) and the residual predictive deviation (RPD) of GA-PLS model are 0.9906, 0.0482, 10.2929. The results show that the GA can reduce the number of modeling variables while ensuring the improvement of model prediction accuracy and robustness.