Fig. 2From: Fast dimension reduction and integrative clustering of multi-omics data using low-rank approximation: application to cancer molecular classificationPerformance of LRAcluster. a the classification accuracy and silhouette value against the dimension of the reduced subspace (the cluster number is set as three) on the three cancer-type testing dataset. b Time consumption of LRAcluster and iCluster+. The number behind the method’s name is the dimension of the latent subspace. iCluster + represents the method that do not tune the penalty parameter. iCluster.tune represents the method tuning the penalty parameter. c and d the dynamic changes of the explained variance and penalty parameter μ as the algorithm iteratesBack to article page