TOWARD FINDING THE BEST MACHINE LEARNING CLASSIFIER FOR LIBS-BASED TISSUE DIFFERENTIATION