Improved Multivariate Calibration Based on Least Square Support Vector Machines

This paper addressed multivariate calibration based on least square support vector machines (LS-SVM) regression to provide a powerful model for machine learning and data mining. LS-SVM technique have the advantages to provide the capability of learning a high dimensional feature with fewer training data, and to decrease the computational complexity for requiring only solving a set of linear equation instead a quadratic programming problem. Experimental results showed the LS-SVM method to be successful for simultaneous multicomponent determination even where there was severe overlap of spectra. It is found that the LS-SVM method is more efficient and accurate than the conventional PLS method.