Consistent bipartite graph co-partitioning for star-structured high-order heterogeneous data co-clustering. She has held postdoctoral or visiting positions at Stanford University, Microsoft.
Biclustering algorithms for biological data analysis: The planar k-means problem is NP-hard. Clustering via matrix powering. Shuchi Chawla, M.
NSF Collaborative Research. Manuscript, accessible at http: Enhanced biclustering on expression data. As future work, we plan to consider several generalizations of Orienteering.
With Shuchi Chawla and Denis Nekipelov. Capturing subspace correlation in a large data set. We give constant factor approximations for various Correlation Clustering objectives for the special case when the similarity measure only classifies pairs as similar or dissimilar, but does not specify the degree of similarity or dissimilarity.
Machine Learning, 75 2: A generalized maximum entropy approach to Bregman co-clustering and matrix approximation. One way of formulating this problem is to maximize the total value of packages that can be delivered in a certain fixed amount of time.