Scatter chart of averaged nRSS index versus number of clusters. The data set is the CAGE tag cluster expression data set from FANTOM3. The number of clusters was estimated by the 10-fold cross validation method. nRSS(e): The Normalized residual sum of squares results calculated by using the estimation dataset. nRSS(v): Result of the validation dataset: a): The value of the number of clusters at which the nRSS(v) reaches the minimum.