Fig. 1From: On Predicting lung cancer subtypes using ‘omic’ data from tumor and tumor-adjacent histologically-normal tissueCross-validation (10-folds) experimental design for a particular classification task, using feature selection and discretization. There are three outcomes: a simple naïve Bayesian model with its test evaluation; clustering of samples based on selected genes; and gene enrichment analysis. Algorithms: ReliefF, Limma, minimum description length principle cut (MDLPC). Evaluation: area under the receiver operating characteristic (AUC), 95 % confidence interval (CI), and Brier Skill Score (BSS)Back to article page