Mehdi Karimi

About Me

I am an Assistant Professor at the Department of Mathematics, Illinois State University. I hold dual PhDs in mathematics and electrical engineering and am passionate about solving real-world challenges through optimization, machine learning, and data-driven decision-making. With expertise at the intersection of theory and practice, I design and implement advanced algorithms for high-impact applications in operations research, power systems, and engineering analytics.

Previously, I was a Visiting Assistant Professor at Florida Tech, a Research Data Scientist at Cerebri AI, and a Postdoctoral Fellow at the University of Waterloo, working with Dr. Levent Tuncel. My research interests include designing optimization algorithms, developing software packages like DDS, and applying machine learning and deep learning techniques to solve complex engineering problems. You can read more about my research here.

I enjoy teaching and exploring innovative learning methods. I completed the Fundamentals of University Teaching and Certificate of University Teaching programs, focusing on peer-instruction in mathematics.

News

March 2024: Awarded a 2-year NSF grant titled "Intelligent Modeling and Parameter Selection in Distributed Optimization for Power Networks."
Jan 2024: Released Domain-Driven Solver (DDS) 2.2, with major updates for quantum relative entropy programming and scalability.

Education

Ph.D. in Combinatorics and Optimization, University of Waterloo, Canada (2013-2017)
Certificate in University Teaching, University of Waterloo, Canada (2013-2014)
M.Sc. in Combinatorics and Optimization, University of Waterloo, Canada (2011-2012)
Ph.D. and M.Sc. in Electrical Engineering, Sharif University of Technology, Iran (2004-2010)
B.Sc. in Electrical Engineering, Isfahan University of Technology, Iran (2000-2004)

Research Interests

Optimization and mathematical programming
Optimization software and applications
Interplay of optimization and machine learning
Hyperbolic programming and sum-of-squares techniques
Optimization in power systems, communications, and smart grids
Convex analysis and interior-point methods

Selected Projects

Distributed Optimization for Power Networks (NSF)

Leading an NSF-funded project involving 4 graduate students to develop decentralized optimization algorithms and graph-based partitioning techniques for modern power grids.

Healthcare Optimization (OSF HealthCare)

Collaborating with 3 faculty and 6 students to predict Medicare claim counts using SDOH data with ML/DL models, addressing challenges in missing data and feature integration.

Maintenance Optimization

Designing reinforcement learning frameworks for predictive maintenance of large-scale infrastructure.

Cybersecurity Risk Models

Developing risk assessment models for cyber-physical vulnerabilities in smart grids and EV charging infrastructure, integrating optimization and statistical learning techniques.