Conference Program Home
  My Program

All Times EDT

Abstract Details

Activity Number: 379 - Single and Multi-Object Regression and Clustering with Applications in Neuro-Imaging Data
Type: Topic Contributed
Date/Time: Wednesday, August 10, 2022 : 8:30 AM to 10:20 AM
Sponsor: Mental Health Statistics Section
Abstract #322709
Title: Biclustering with Feature Selection
Author(s): Eric Chi* and Jason Xu and Saptarshi Chakraborty and Debolina Paul
Companies: Rice University and Duke University and University of California, Berkeley and Stanford
Keywords: Variable Selection; Optimization; Penalized Regression; Co-Clustering

Convex clustering has recently gained popularity due to its attractive theoretical and computational properties. Though it confers many advantages over traditional clustering methods, its merits become limited in the face of high-dimensional data due to numerical instability. To address these issues, biconvex clustering has been proposed recently to simultaneously optimize the feature weights as well as the centroids. In the biclustering problem, we seek to simultaneously group observations and features. For high-throughput bioinformatics data, we often also wish to perform feature selection. In this talk, we propose a novel biconvex algorithm to tackle the biclustering problem while simultaneously performing feature selection. Called `biconvex biclustering', the proposed method performs by selecting proper subsets in both the groups as well as the features throughout the clustering task. We demonstrate our method’s utility for exploring single cell RNA sequencing data.

Authors who are presenting talks have a * after their name.

Back to the full JSM 2022 program