SIAM MDS20 MS13 Advances in Subspace Learning and Clustering Mini-Symposium
Abstract.: State-of-the-art subspace clustering methods are based on expressing each data point as a linear combination of other data points while regularizing the matrix of coefficients with sparse,or low-rank regularizations. Sparse regularization is guaranteed to give a subspace-preserving affinity (i.e., there are no connections between points from different subspaces) under broad conditions, but the clusters may not be connected. Low-rank and regularization often improve connectivity, but give a subspace-preserving affinity only for independent subspaces. Moreover, the applicability of such methods to large datasets has been limited due to time complexity and/or memory footprint issues.
In this talk, I present a novel subspace clustering approach based on the elastic net regularizer (a mixture of the sparse andregularizers). Our geometric analysis shows that this method maintains subspace-preserving property (due to sparse regularization) while enhances connectedness (due to regularization), therefore achieves better clustering performance. We also present two provably correct and scalable active support algorithms for solving the elastic net regularized optimization problem. Experiments on synthetic data verify our theoretical analysis, and show that the proposed methods efficiently handle datasets with 1M data points. In addition, our methods achieve % clustering accuracy on images in the MNIST dataset in about 1 hour.
- Chong You, University of California, Berkeley, U.S., email@example.com
Join Zoom Meeting
Meeting ID: 982 2294 6324