Vladimir Dvorkin

My work focuses on decision-making models and algorithms for enhancing planning, operation, and pricing in electric power systems with substantial renewable energy integration and rising AI-driven electricity demand. This involves developing prescriptive analytics—such as decision-focused and physics-informed machine learning models—to manage grid uncertainties, building large-scale equilibrium models to analyze interactions among AI users in the grid, and investigating possible biases arising from machine learning applications in electricity markets. To enable this work on proprietary and sensitive datasets, I develop privacy-preserving algorithms for energy analytics and data synthesis from real-world infrastructure systems using mathematical optimization and differential privacy.