The attached files contain two datasets termed GCM and Acute Leukemia datasets. They have been used in our submitted paper termed 'Semi-supervised Nonnegative Matrix Factorization for Cancer Classification'. The file names imply its datasets. The details are listed as follows: GCM_Test.res --- The testing gene expression data of the GCM dataset GCM_Test.cls --- The ground-truth classes of the testing samples of the GCM dataset GCM_Training.res --- The training gene expression data of the GCM dataset GCM_Training.cls --- The ground-truth classes of the training samples of the GCM dataset GCM.mat --- The .mat file contains the training and testing data and their classes information. The GCM dataset were downloaded from the website http://www.broadinstitute.org/cgi-bin/cancer/datasets.cgi. Leukemia dataset Acute_Leukemia_feature.txt --- The feature names Acute_Leukemia_sampleLabel.txt --- The ground-truth classes of all samples Acute_Leukemia_sampleName.txt --- The sample names Acute_Leukemia1.txt --- The file is the gene expression of the first 1048 samples Acute_Leukemia2.txt --- The file is the gene expression of the subsequent 1048 samples Acute_Leukemia.mat --- The .mat file contains gene expression data of all the samples and their classes. The Leukemia dataset were obtained from the website : http://www.ncbi.nlm.nih.gov/geo/query/acc.cgi?acc=GSE13159