Additional file 10 of Common pathways and functional profiles reveal underlying patterns in Breast, Kidney and Lung cancers

Additional file 10: Table S5. Metrics results from the algorithm comparison through K-Fold Cross-Validation to classify three types of tumors. Description: The comparison has been made by grouping the tumor samples of the three types of cancer (BRCA, LUAD, KIRC), taking in account either all significant paths or only the top ten most significant ones. The algorithms used were K-Nearest-Neighbor (KNN), Decision Tree (CART), Naive-Bayes Classificator (NB), K-Means clustering (KMN) and Random Forest (RF). The metrics analysed take account of the mean Accuracy of the model testing across all the possible training and test combinations data splits resulting from the K-Fold Cross-Validation, along with its associated Standard Deviation.