A Bicluster-Based Bayesian Principal Component Analysis Method for Microarray Missing Value Estimation

F. Meng, C. Cai, and H. Yan

The Bayesian Principal Component Analysis (BPCA) model is integrated into the biclustering framework to estimate missing values in microarray data. The local similarity structure is fully exploited in biclusters, which overcomes the drawback of BPCA that only global feature of the entire matrix is utilized.

Read more at IEEE Xplore

Tags: Bayesian principal component analysis (BPCA), Biclustering, Microarray missing value estimation