Ehsan Abbasnejad

Papers

Total Citations

H-Index

About

Ehsan Abbasnejad is a leading researcher in probabilistic machine learning and artificial intelligence, with a core focus on advancing inference and decision-making under uncertainty. His most significant contribution is the development of **Symbolic Variable Elimination (SVE)** for discrete and continuous graphical models, a breakthrough method that enables exact, closed-form probabilistic inference even when joint distributions are non-Gaussian. This work, published in 2021 and garnering 30 citations, addresses a long-standing gap in the field by extending classical variable elimination to handle complex, real-world continuous systems without resorting to approximations. Abbasnejad is also recognized for his work on **Loss-Calibrated Monte Carlo Action Selection** (2015), which integrates Bayesian decision theory into action selection to hedge against catastrophic outcomes in high-stakes domains like robotics and plant control. By bridging rigorous probabilistic reasoning with practical decision-making, his research has profound implications for autonomous systems, robotics, and safety-critical AI. His work stands out for its theoretical elegance and direct applicability, making him a key figure in the push toward more reliable and robust intelligent systems.