Salar Fattahi helping high school student solve math problems

Salar Fattahi receives NSF CAREER Award for work in nonconvex problems in machine learning

Salar Fattahi discusses his award-winning CAREER proposal which aims to close the current gap between optimization and statistical learning.

Salar Fattahi

Salar Fattahi an assistant professor at the University of Michigan Industrial and Operations Engineering Department (U-M IOE) has received the National Science Foundation (NSF) Faculty Early Career Development (CAREER) Award. His proposed research titled “Blessing of Nonconvexity in Machine Learning – Landscape Analysis and Efficient Algorithms” will explore how to uncover the distinct structures of the nonconvex problems in machine learning that will make them tractable, reliable and widely available in a variety of machine learning applications.

As one of NSF’s most prestigious awards, the CAREER award supports early-career faculty who have the potential to serve as academic role models in research and education and to advance the mission of their department. 

“I feel incredibly happy to have been honored with this prestigious award,” said Fattahi. “I would like to thank my exceptional students and colleagues within and outside IOE, whose support and contributions made this achievement possible.”

Fattahi’s CAREER proposal aims to close the current gap between optimization and statistical learning and will challenge the conventional paradigms that evaluate the performance of optimization algorithms solely based on their ability to find global optima. He will do this by conducting a formal systematic analysis of the optimization landscape of nonconvex models around the true solutions and will then design reliable and efficient algorithms to solve them in meaningful settings and scales.

Salar Fattahi stands next to a circular table where four people are sitting and writing on sheets of paper as they do math problems.
Salar leads a workshop on Human-AI Interaction for the Early College Alliance High School, hosted by the University of Michigan Industrial and Operations Engineering Department.

“With this award, I strive to develop reliable, efficient, and publicly available computational solvers that are applicable to a wide range of optimization problems arising in machine learning,” said Fattahi. “I envision that the efficiency and reliability of these methods will enable domain experts and practitioners to use them seamlessly in their related fields, even under limited computational resources and energy budgets.”

A part of this project will include the development of a variety of educational programs for K-12 and higher education students. Fattahi plans to create new partnerships with under-resourced schools to help introduce new college opportunities to students from low-income families. Additionally, he will broaden the impact of these programs by making the materials publicly available.

Fattahi joined the U-M IOE as a faculty member in 2020. Fattahi received his BSc degree in Electrical Engineering from the Sharif University of Technology in 2014. He went on to obtain his MSc and PhD degrees in Industrial Engineering and Operations Research from the University of California, Berkeley. While receiving an additional MSc degree in Electrical Engineering from Columbia University. His work has been funded by NSF, the Office of Naval Research, and various U-M grants. 

To learn more about Fattahi’s accomplishments and work visit his website.