NSF Awards SU's Anderson $174K for Machine Learning and Theoretical Computer Science Research

SALISBURY, MD---The National Science Foundation (NSF) is helping a Salisbury University mathematics and computer science professor expand research and student education in machine learning and theoretical computer science.  

Dr. Joseph Anderson received a $174,000 NSF grant to continue independent research on exploiting geometry in robust signal processing and feature extraction.

“I am thrilled for this opportunity,” Anderson said. “I hope with this grant I can build some momentum to further develop a machine learning and theoretical computer science research group within our department.”    

The funding will support student assistant compensation, room, board, and national and international conference travel. There also is funding for advanced computing hardware that will be dedicated to this project, run specialized software and link into the Henson school’s High-Performance Computing Lab (HPC) to leverage that computing power.

“With the new HPC cluster at SU, this can be done a lot faster,” Anderson said. “We can get a much better intuition with the help of computers and then attack the problems with purely mathematical tools.”

The core of the project studies which of several existing problem frameworks and algorithms in machine learning and signal processing can be made tolerant of heavy-tailed noise or adversarial corruption. It will involve significant study of state-of- the-art algorithms, usually with synthetic data, and find where there are gaps in current understanding.

“The essence of this project is to develop tools to relax the assumptions that data is ‘well-behaved’ enough that algorithms work correctly, and enable standard algorithms to function when data is very noisy, or even corrupted by a malicious user,” said Anderson. 

This project was ideal for sophomore mathematics, computer science and data science major Blaine Mason, of Salisbury, who is assisting Anderson throughout the grant.

“I was interested in the opportunity to learn an immense amount of material that, before being approached by Dr. Anderson, I knew nothing about,” said Mason. “Not only material about computational geometry, but working with others and approaching problems that do not have straight-forward answers. One of my favorite things is to shift my perspective on a question, and I feel that the research we are conducting is a great chance to do that.”

This work is more than a summer project for Mason, who says the benefits will span beyond undergraduate study into graduate school and even into his own university courses someday. 

“Having this opportunity and learning these skills, I’m able to prepare for the research I will be doing at a graduate level and learn how to include students in my own research when I become a professor,” he said. 

For more information call 410-543-6030 or visit the SU website.