Strategies for improving performance of evolutionary biclustering algorithm EBIC

Biclustering is a growing in popularity machine learning technique which searches for patterns in subsets of rows and subsets of columns. One of the recent advances in biclustering was the development of EBIC, a multi-GPU method based on evolutionary computation, which was demonstrated to outperform some of the leading methods in the field. In this short paper, we evaluate a couple of potential improvements to the method.