Seminar on Computational Learning and Adaptation


  Constrained Clustering for Improved Pattern Discovery

Sepander Kamvar
Department of Scientific Computing/Computational Mathematics
Stanford University

We present an improved method for clustering in the presence of very limited supervisory information, given as pairwise instance constraints. By allowing instance-level constraints to have space-level inductive implications, we are able to successfully incorporate constraints for a wide range of data set types. Our method greatly improves on the previously studied COP-K-means algorithm, generally requiring less than half as many constraints to achieve a given accuracy on a range of real-world data. We additionally discuss an active learning algorithm which increases the value of constraints even further.



Date: Thursday, October 3

Time: 4:15-5:30PM

Place: Cordura 100


Return to the seminar schedule