Conference Program Home
  My Program

All Times EDT

Abstract Details

Activity Number: 65 - Computational Challenges in Software and Algorithms
Type: Contributed
Date/Time: Sunday, August 7, 2022 : 4:00 PM to 5:50 PM
Sponsor: Section on Statistical Computing
Abstract #323405
Title: CUR Algorithm for Data Analysis
Author(s): Kathryn Linehan* and Radu Balan
Companies: University of Virginia, Biocomplexity Institute, Social and Decision Analytics Division and University of Maryland, Dept of Mathematics & Center for Sci Computation and Math Modeling
Keywords: computational statistics; statistical applications; machine learning; matrix decomposition; data science

Data analysis depends not only on appropriate methods, but also on interpretation of results. We present a CUR matrix decomposition algorithm which provides intuitive interpretation in terms of the original data for applications and could potentially be used as an alternative to principal component analysis. As matrix decompositions are paramount to computational statistics, our CUR algorithm also contributes to this field. Specifically, our algorithm utilizes optimization problems similar to that of lasso regression to determine the matrices C and R.

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

Back to the full JSM 2022 program