Research Experience

Doctoral Researcher, Dept. of Mathematical Sciences, RPI (May 2020 – Present)

• Designed and analyzed a state-of-the-art optimization algorithm, Damped Proximal Augmented Lagrangian Method (DPALM), to improve convergence speed in nonconvex problems.
• Evaluated complexity and performance of DPALM in applications including bias reduction in machine learning and robustness enhancement.
• Solved dynamic optimal transport problems with real-world applications in generative AI using decentralized optimization techniques.
• Implemented algorithms in Python using MPI and Numpy for distributed computation.

Research Interests

• Constrained Optimization Algorithms
• First-order Methods & Nonconvex Optimization
• Dynamic Optimal Transport Problems
• Distributed Optimization
• Convergence Analysis