Ssignment of Acute Myelogenous Leukemia (AML) and Acute Lymphoblastic Leukemia (ALL)Kmeans (K) ALL__B.cell ALL__B.cell ALL__B.cell ALL__B.cell ALL__B.cell ALL__B.cell ALL__B.cell ALL__B.cell ALL__B.cell ALL__B.cell ALL__B.cell ALL__B.cell ALL__B.cell ALL__B.cell ALL__B.cell ALL__B.cell ALL__B.cell ALL_R_B.cell ALL_R_B.cell ALL__T.cell ALL__T.cell ALL__T.cell ALL__T.cell ALL__T.cell ALL__T.cell ALL__T.cell ALL__T.cell AML_ AML_ AML_ AML_ AML_ AML_ AML_ AML_ AML_ AML_ AML_ Error Count L L L L L L L L L L L L L L L L L L L L L L L L L L L L L M M L L L M L L M SVD (K) B T M B B B M T M T B M M B M T M M B T T T T M T T M M M M T B B M T B M T L L L L L L L L L L L L L L L L L L L L M M M L M L M L L L L L L L L L L LPCA (K) M B M B M M M M B B M M M B M B M M B B B T T T B B B M B M M M M M M M M M L L L L L M L L L L L L L L L L M L L L L L L L L L L M L M M M M M M M M MICA (K) B B B B B B B B B B B B B B B B B B B B B T T T B B T M B M M M M M M M M M L L L L L M L L L L L L L L L L M L L L L L L L L L L M M M M M M M M M M MNMF (K) B B B B B M B B B T B B B B B B M B B T T T T T T T T M M M M M M M M M M MSNMF (K) B B B B B B B B B T B B B B B B B B B T T T T T T T T M B M M M M M M M M MBSNMF (K) B B B B B B B B B B B B B B B B B B B T T T T T T T T M B M M M M M M M M MVoting (K) B B B B B B B B B B B B B B B B B B B T T T T T T T T M B M M M M M M M M M(K)(K) M L M M M M L L M L M M L M L L L L M L L L L L L L L L L L L M M L L M L L(K)(K)(K)(K) L L L L L L L L L L L L L L L L L L L L L L L L L L L M L M M M M M M M M M(K) L L L L L L L L L L L L L L L L L L L L L L L L L L L M L M M M M M M M M M(K) L L L L L L L L L L L L L L L L L L L L L L L L L L L M L M M M M M M M M MB B B B B B B B B B B B B B B B B B B T T T T T T T T B B M M B B B M B B MClass Assignment of Acute Myelogenous Leukemia (AML) and Acute Lymphoblastic Leukemia (ALL) at K and K. SVD: singular worth decomposition, PCA: principal component evaluation, ICA: independent element evaluation, NMF: non-negative matrix factorization, SNMF: sparse non-negative matrix factorization, BSNMF: bi-directional non-negative matrix factorization, Voting: Voting class L: ALL, M: AML, B: ALL_B cell, T: ALL_T cell Bold-faced: misclassified samplesevaluation study for the solutions has not been reported. Hence, we evaluated orthogonal (i.e. PCA, SVD), GSK 137647 chemical information non-orthogonal (i.e. ICA, NMF and SNMF) MFs and also a conventional clustering algorithm (i.e. K-means) employing seven clustering-quality (i.e. homogeneity, separation, Dunn index, average silhouette width, Pearson correlation 
of cophenetic distance, Hubert correlation of cophenetic distance and the GAP statistic) and two prediction-accuracy measures (i.e. the adjusted Rand index and prediction accuracy) applying to 5 published datasets. We also incorporated an improving non-orthogonal MFs, BSNMF inside the evaluation study. Because of this, we observed that clustering quality and prediction-accuracy indices applying non-orthogonalMFs are better than these of orthogonal MFs and Kmeans. In respect to benefits from Homogeneity, separation, Dunn index, average silhouette width and Hubert correlation of cophenetic distance, non-orthogonal MFs had greater worth than those of orthogonal MFs and Kmeans. The GAP statistic was reduce DM1 pubmed ID:http://www.ncbi.nlm.nih.gov/pubmed/19387489?dopt=Abstract for non-orthogonal MFs than for orthogonal MFs and K-means. When we tested predictive accuracy for the 3 datasets with known class labels, we also observed improved overall performance for non-orthogonal MFs than for the rest. We also investig.Ssignment of Acute Myelogenous Leukemia (AML) and Acute Lymphoblastic Leukemia (ALL)Kmeans (K) ALL__B.cell ALL__B.cell ALL__B.cell ALL__B.cell ALL__B.cell ALL__B.cell ALL__B.cell ALL__B.cell ALL__B.cell ALL__B.cell ALL__B.cell ALL__B.cell ALL__B.cell ALL__B.cell ALL__B.cell ALL__B.cell ALL__B.cell ALL_R_B.cell ALL_R_B.cell ALL__T.cell ALL__T.cell ALL__T.cell ALL__T.cell ALL__T.cell ALL__T.cell ALL__T.cell ALL__T.cell AML_ AML_ AML_ AML_ AML_ AML_ AML_ AML_ AML_ AML_ AML_ Error Count L L L L L L L L L L L L L L L L L L L L L L L L L L L L L M M L L L M L L M SVD (K) B T M B B B M T M T B M M B M T M M B T T T T M T T M M M M T B B M T B M T L L L L L L L L L L L L L L L L L L L L M M M L M L M L L L L L L L L L L LPCA (K) M B M B M M M M B B M M M B M B M M B B B T T T B B B M B M M M M M M M M M L L L L L M L L L L L L L L L L M L L L L L L L L L L M L M M M M M M M M MICA (K) B B B B B B B B B B B B B B B B B B B B B T T T B B T M B M M M M M M M M M L L L L L M L L L L L L L L L L M L L L L L L L L L L M M M M M M M M M M MNMF (K) B B B B B M B B B T B B B B B B M B B T T T T T T T T M M M M M M M M M M MSNMF (K) B B B B B 
B B B B T B B B B B B B B B T T T T T T T T M B M M M M M M M M MBSNMF (K) B B B B B B B B B B B B B B B B B B B T T T T T T T T M B M M M M M M M M MVoting (K) B B B B B B B B B B B B B B B B B B B T T T T T T T T M B M M M M M M M M M(K)(K) M L M M M M L L M L M M L M L L L L M L L L L L L L L L L L L M M L L M L L(K)(K)(K)(K) L L L L L L L L L L L L L L L L L L L L L L L L L L L M L M M M M M M M M M(K) L L L L L L L L L L L L L L L L L L L L L L L L L L L M L M M M M M M M M M(K) L L L L L L L L L L L L L L L L L L L L L L L L L L L M L M M M M M M M M MB B B B B B B B B B B B B B B B B B B T T T T T T T T B B M M B B B M B B MClass Assignment of Acute Myelogenous Leukemia (AML) and Acute Lymphoblastic Leukemia (ALL) at K and K. SVD: singular value decomposition, PCA: principal element evaluation, ICA: independent component analysis, NMF: non-negative matrix factorization, SNMF: sparse non-negative matrix factorization, BSNMF: bi-directional non-negative matrix factorization, Voting: Voting class L: ALL, M: AML, B: ALL_B cell, T: ALL_T cell Bold-faced: misclassified samplesevaluation study for the solutions has not been reported. Hence, we evaluated orthogonal (i.e. PCA, SVD), non-orthogonal (i.e. ICA, NMF and SNMF) MFs and a regular clustering algorithm (i.e. K-means) using seven clustering-quality (i.e. homogeneity, separation, Dunn index, typical silhouette width, Pearson correlation of cophenetic distance, Hubert correlation of cophenetic distance and also the GAP statistic) and two prediction-accuracy measures (i.e. the adjusted Rand index and prediction accuracy) applying to 5 published datasets. We also integrated an improving non-orthogonal MFs, BSNMF in the evaluation study. Because of this, we observed that clustering high quality and prediction-accuracy indices applying non-orthogonalMFs are better than these of orthogonal MFs and Kmeans. In respect to results from Homogeneity, separation, Dunn index, average silhouette width and Hubert correlation of cophenetic distance, non-orthogonal MFs had higher value than these of orthogonal MFs and Kmeans. The GAP statistic was decrease PubMed ID:http://www.ncbi.nlm.nih.gov/pubmed/19387489?dopt=Abstract for non-orthogonal MFs than for orthogonal MFs and K-means. When we tested predictive accuracy for the three datasets with recognized class labels, we also observed superior overall performance for non-orthogonal MFs than for the rest. We also investig.