Rectified Factor Networks for Biclustering

Djork-Arné Clevert, Thomas Unterthiner, Sepp Hochreiter

Nov 02, 2016 (modified: Nov 02, 2016) ICLR 2017 conference submission readers: everyone
  • Abstract: Biclustering is evolving into one of the major tools for analyzing large datasets given as matrix of samples times features. Biclustering has several noteworthy applications and has been successfully applied in life sciences and e-commerce for drug design and recommender systems, respectively. FABIA is one of the most successful biclustering methods and is used by companies like Bayer, Janssen, or Zalando. FABIA is a generative model that represents each bicluster by two sparse membership vectors: one for the samples and one for the features. However, FABIA is restricted to about 20 code units because of the high computational complexity of computing the posterior. Furthermore, code units are sometimes insufficiently decorrelated. Sample membership is difficult to determine because vectors do not have exact zero entries and can have both large positive and large negative values. We propose to use the recently introduced unsupervised Deep Learning approach Rectified Factor Networks (RFNs) to overcome the drawbacks of existing biclustering methods. RFNs efficiently construct very sparse, non-linear, high-dimensional representations of the input via their posterior means. RFN learning is a generalized alternating minimization algorithm based on the posterior regularization method which enforces non-negative and normalized posterior means. Each code unit represents a bicluster, where samples for which the code unit is active belong to the bicluster and features that have activating weights to the code unit belong to the bicluster. On 400 benchmark datasets with artificially implanted biclusters, RFN significantly outperformed 13 other biclustering competitors including FABIA. In biclustering experiments on three gene expression datasets with known clusters that were determined by separate measurements, RFN biclustering was two times significantly better than the other 13 methods and once on second place. On data of the 1000 Genomes Project, RFN could identify DNA segments which indicate, that interbreeding with other hominins starting already before ancestors of modern humans left Africa.
  • Conflicts:
  • Keywords: Deep learning, Unsupervised Learning, Applications