BIDS Faculty Affiliate Michael Mahoney and former BIDS Fellows Michał Dereziński and Rajiv Khanna of Berkeley’s Foundations of Data Analysis Institute (FODA) have been awarded a “NeurIPS 2020 Best Paper Award” for their contribution, Improved Guarantees and a Multiple-Descent Curve for Column Subset Selection and the Nyström Method.
By expanding on approximation techniques that have been widely used in machine learning, this paper expands our understanding of the foundations of data for reproducible and interpretable data analysis. It is expected to have substantial impacts and yield new insights into a variety of applications including kernel methods, feature selection, and the double-descent behavior of neural networks.
The paper will be presented on Wednesday, December 9, at 6:00 PM PT as part of the “Learning Theory” track. This year’s Neural Information Processing Systems Conference will be presented online and feature a variety of presentations, posters, tutorials and awards.
Announcing the NeurIPS 2020 award recipients
December 7, 2020 | Neural Information Processing Systems Conference
Hsuan-Tien Lin, Maria Florina Balcan, Raia Hadsell and Marc’Aurelio Ranzato
Improved Guarantees and a Multiple-Descent Curve for Column Subset Selection and the Nyström Method
November 23, 20202 | arXiv.org
Michał Dereziński, Rajiv Khanna, and Michael W. Mahoney