This Saturday, I will talk about the paper "Biclustering with heterogeneous variance". Biclustering is a fundamental problem in bioinformatics. The authors proposed an statistical approach based on SVD which captures both mean and variance structure. The authors
demonstrated the effectiveness of their approach on cancer mythelation data and lung cancer gene expression data.