Boxplots of testing accuracies of the LPPO with four gene selection methods using two different classifiers (NBC, NMSC) compared to varSelRF for six data sets. RF is the final classifier. All six data sets demonstrate that varSelRF accuracies are lower than our proposed feature selection and optimization algorithm with the same RF classifier.
Liu et al. BMC Genomics 2011 12(Suppl 5):S1 doi:10.1186/1471-2164-12-S5-S1