Figure 2

The analysis of the three-fold cross validation performance of the algorithm on the Leukemia dataset. The dotted lines indicate the performance values in individual tests. The solid lines indicate the average values; and the dotted lines indicate one standard deviation from the averages. The X-axis represents the number of genes in S _i . (a) The classification accuracy θ _test on the test samples. (b) The classification accuracy θ _train on the training samples. (c) The p-values ρ _S (i) based on $\overline{d}(A_i,{{\Delta }^{\prime }}_1)$. (d) The p-values ρ _S (i) based on $\overline{d}(A_i,{{\Delta }^{\prime }}_2)$. (e) The p-value ρ _all (i) based on $\widehat{\delta }$(A _i ). (f) The average Silhouette width $\overline{\omega }$(A _i ) of active clusters in A _i .