Table 1.

Summary of classification accuracies both for our approach and Random Forests.

Method	Number of ratios in the model	Correct classification rate on the learning dataset	2-fold cross-validation average correct classification rate	Corect classification rate on the validation dataset
GA + LDA	2	86.7%	81.7%	86.7%
RF	5112	100%	67.3%	70%
RF	6	100%	86.4%	76.7
RF	2	100%	79.4%	50%