News Release

Unifying statistics, computer science, and applied mathematics

Lehigh, Stony Brook, and Northwestern team with NSF to advance machine learning

Grant and Award Announcement

Lehigh University

Katya Scheinberg, Lehigh University

image: Professor Katya Scheinberg, Industrial and Systems Engineering at Lehigh University, has been working on the intersection of optimization and machine learning for over 15 years. view more 

Credit: Christa Neu / Lehigh University

The National Science Foundation (NSF) has announced its support of a Lehigh University-led research team that will advance machine learning by tying together techniques drawn from the fields of statistics, computer science, and applied mathematics.

The three-year, $1.5-million grant is part of the NSF's $17.7 million in funding for 12 Transdisciplinary Research in Principles of Data Science (TRIPODS) projects, which will bring together statisticians, theoretical computer scientists, and mathematicians to advance the foundations of data science. Conducted at 14 institutions in 11 states, these projects will promote long-term research and training activities in data science that transcend disciplinary boundaries.

"Data is accelerating the pace of scientific discovery and innovation," said Jim Kurose, NSF Assistant Director for Computer and Information Science and Engineering, in the NSF's August 24 announcement. "These new TRIPODS projects will help build the theoretical foundations of data science that will enable continued data-driven discovery and breakthroughs across all fields of science and engineering."

The Lehigh project is led by three members of the University's Industrial and Systems Engineering department: principal investigator Katya Scheinberg and co-investigators Frank E. Curtis and Martin Takáč. The project is a collaborative effort with Han Liu of Northwestern University and Francesco Orabona of SUNY-Stony Brook.

"Progress in the field of machine learning requires close collaboration among optimization experts, learning theorists, and statisticians," says Scheinberg, Lehigh's Harvey E. Wagner Endowed Chair Professor of Industrial Engineering. "Machine learning draws so heavily from these areas, yet the communities supporting research in each have tended to operate separately -- each with its own vocabulary and platforms for publishing state-of-the-art research. With an emphasis on deep learning, our project aims to build bridges and foster intercollegiate and interdisciplinary collaboration among these communities."

According to the NSF, all of the TRIPODS awards will enable data-driven discovery through major investments in state-of-the-art mathematical and statistical tools, better data mining and machine learning approaches, enhanced visualization capabilities, and more. These awards will build upon NSF's long history of investments in foundational research, contributing key advances to the emerging data science discipline, and supporting researchers to develop innovative educational pathways to train the next generation of data scientists.

The Lehigh "tripod"

Working with Orabona, a theoretical computer scientist specializing in learning theory at SUNY-Stony Brook, and Liu, an expert in statistics and machine intelligence at Northwestern, the three-pronged Lehigh team hopes to be a beacon for those who wish to learn and help advance the state-of-the-art in machine learning throughout the northeastern and midwestern U.S. The team's TRIPODS project is built upon the ongoing efforts of Lehigh's Optimization and Machine Learning (OptML) research group, founded in 2014. Research conducted by the members of the OptML group focuses on the design, analysis, and implementation of numerical methods for solving large-scale optimization problems arising in machine learning applications.

Each Lehigh researcher involved in the TRIPODS project brings to the team unique expertise in applied mathematics, particularly in topics related to mathematical optimization:

  • Professor Katya Scheinberg has been working on the intersection of optimization and machine learning for more than 15 years. She is especially well known for her work on kernel support vector machines, a widely applicable and powerful data science tool. Scheinberg earned her undergraduate degree in Operations Research from the Lomonosov Moscow State University and her Ph.D. in Industrial Engineering and Operations Research from Columbia University. Prior to her appointment at Lehigh, she served as a researcher at IBM's famed T.J. Watson Research Center for more than a decade, working on various applied and theoretical problems in optimization. She is currently the Editor-in-Chief of the SIAM-MOS Series on Optimization and an Associate Editor of SIAM Journal on Optimization and Mathematical Programming. She has other research projects supported by the U.S. Air Force, DARPA, and Yahoo, as well as the NSF.
  • Associate Professor Frank E. Curtis brings to the team widely recognized expertise in solving nonconvex and nonsmooth optimization problems. This expertise is essential in the context of deep learning, a main focus of the project. He earned his Bachelor's degree in Mathematics and Computer Science (double major) from the College of William and Mary and his Master's and Ph.D. degrees from Northwestern University in Industrial Engineering and Management Science. Prior to joining Lehigh in 2009, he spent two years as a postdoctoral researcher at the Courant Institute of Mathematical Sciences. Curtis' research focuses on the design, analysis, and implementation of numerical methods for solving large-scale nonlinear optimization problems. He received an Early Career Award from the Advanced Scientific Computing Research program of the U.S. Department of Energy, and has received funding from various NSF programs. He served as the Vice Chair for Nonlinear Programming for the INFORMS Optimization Society from 2010 until 2012, is currently an Associate Editor for SIAM Journal on Optimization and Mathematical Programming Computation, and is very active in professional societies and groups related to mathematical optimization.
  • Assistant Professor Martin Takáč brings to the team widely recognized expertise in designing efficient algorithms for solving large-scale optimization problems, especially those associated with deep learning, including coordinate descent and stochastic gradient-descent-type algorithms. Takáč has extensive experience in high performance computing, and has performed research using some of the world's largest supercomputers. Takáč received his B.S. and M.S. degrees in Mathematics from Comenius University, Slovakia, and his Ph.D. degree in Mathematics from The University of Edinburgh, United Kingdom. He received several awards during this period, including the Best Ph.D. Dissertation Award by the OR Society (2014), the Leslie Fox Prize (2nd Prize; 2013) from the Institute for Mathematics and its Applications, and the INFORMS Computing Society Best Student Paper Award (runner up; 2012). Since joining the faculty of Lehigh's Department of Industrial and Systems Engineering in 2014, he has pursued research in such areas as design, analysis and application of algorithms for machine learning, optimization, high-performance computing, operations research, and energy systems. He is an affiliated faculty member of Lehigh's Cognitive Science Program.


Disclaimer: AAAS and EurekAlert! are not responsible for the accuracy of news releases posted to EurekAlert! by contributing institutions or for the use of any information through the EurekAlert system.