Lewis-Sigler Institute for Integrative Genomics, Princeton University
Download the paper here
Download the supplementary material here
Run the Iclust algorithm with your data here
In an age of increasingly large data sets, investigators in many different disciplines have turned to clustering as a tool for data analysis and exploration. Existing clustering methods, however, typically employ nontrivial assumptions. Here we reformulate the clustering problem from an information theoretic perspective which avoids many of these assumptions. Our formulation obviates the need for defining a cluster "prototype", does not require an a priori similarity metric, is invariant to changes in the representation of the data, and naturally captures non--linear relations. We apply this approach to different domains and find that it consistently produces clusters that are more coherent than those extracted by existing algorithms. Finally, our approach provides a way of clustering based on collective notions of similarity rather than the traditional pairwise measures.
Acknowledgements: Gasper Tkacik was in part supported by Burroughs-Wellcome Graduate Training Program in Biological Dynamics