Sanjeev Arora – Communications of the ACM

Research and Advances Mar 1 2021

Technical Perspective: Why Don’t Today’s Deep Nets Overfit to Their Training Data?

"Understanding Deep Learning (Still) Requires Rethinking Generalization," Chiyuan Zhang, et al., brings a fundamental new theoretical challenge: Why don't today's deep nets overfit to training data?

Artificial Intelligence and Machine Learning

Author Archives

Research and Advances Apr 1 2018

Learning Topic Models – Provably and Efficiently

This article shows that some new theoretical algorithms that have provable guarantees can be adapted to yield highly practical tools for topic modeling.

Artificial Intelligence and Machine Learning

Research and Advances May 1 2011

Computational Complexity and Information Asymmetry in Financial Products

Securitization of cash flows using financial derivatives transformed the financial industry over the last three decades. Derivatives have attracted criticism, but others say problems with derivatives would disappear with use of more accurate financial models, more vigilance by buyers and better governmental oversight.

Artificial Intelligence and Machine Learning

Research and Advances Oct 1 2008

Geometry, Flows, and Graph-Partitioning Algorithms

"Graph partitioning" refers to a family of computational problems in which the vertices of a graph have to be partitioned into two (or more) large pieces while minimizing the number of the edges that cross the cut. The goal of this paper is to survey an interesting combination of techniques that have recently led to progress on graph partioning problems.

Systems and Networking

Opinion Aug 1 2005

Is the Thrill Gone?

Computer science is nothing less than a new way of thinking; explaining it to the wider world is critical to the future of the field.

Computing Applications

Shape the Future of Computing

ACM encourages its members to take a direct hand in shaping the future of the association. There are more ways than ever to get involved.

Get Involved