Embedded Files
I develop interpretable AI methods for scientific research, focusing on optimization and simulation-based approaches that create transparent mathematical and computational models of physical and complex systems. My work enables testing hypotheses and understanding causal mechanisms in ways that traditional black-box models cannot.
My research centers on improving sequential decision making under uncertainty, building embodied agents for decision making, and addressing the technical challenges of large-scale data-driven systems. I'm particularly passionate about simulation-based inference, geometric learning, and developing hybrid models that combine the strengths of deep learning with traditional mathematical approaches.
Throughout my career, I've tackled research problems across diverse domains from defense to supply chains. My long-term vision is to bridge deep learning with traditional mathematical models of complex and collective systems—those involving many interacting components like cells, active particles, and agents. These systems fascinate me because they operate far from equilibrium, exhibit stochastic behavior, exist in high dimensions, and demonstrate remarkable emergent properties.
I'm working toward discovering new rules governing adaptive and complex systems and integrating these insights into active learning and experimental design. This involves searching through high-dimensional problem spaces using emulation, simulation-based surrogate methods, and curriculum learning to achieve faster optimization convergence.
1. Combinatorial Optimization
1. Combinatorial Optimization
In the realm of combinatorial optimization, my work is primarily focused on developing and implementing robust mathematical models and computational algorithms. The essence of these efforts is to address the inherently complex nature of planning and scheduling problems that manifest across diverse business and service sectors. Such problems include scheduling and routing of on-demand services , airline scheduling and routing, sports scheduling, physician and nurse scheduling, delivery routing, crew/gang scheduling, supply chain network design, and many others.
2. Simulation Modeling and Digital Twins
2. Simulation Modeling and Digital Twins
A significant portion of my work is dedicated to simulation modeling and the creation of digital twins, which are virtual replicas of physical systems. This technology allows for real-time monitoring and optimization of operations, which facilitates proactive decision-making and enhances operational efficiency.
3. Reinforcement Learning and Intelligent Control
3. Reinforcement Learning and Intelligent Control
My research in reinforcement learning (RL) explores the potential of algorithms that learn optimal actions through trial and error interactions with an environment, aiming to maximize some notion of cumulative reward. In contrast to the more common machine learning tasks of predicting and classifying, RL is used for planning and control tasks.
4. Logistics and Supply Chain Management
4. Logistics and Supply Chain Management
Supply chains and logistical services constitute some of the most essential pillars of our modern economy. My work focus on developing algorithms for better inventory management to reduce costs and improve service levels, and designing models for logistics optimization to ensure timely delivery of goods and services, critical for customer satisfaction and business success.
I'm excited about the potential impact of my research, and I'm always eager to collaborate with others who share my passion for model-based and data-driven mathematical and computational techniques. Feel free to reach out if you have any questions or if you're interested in discussing potential research collaborations!
Page updated
Google Sites
Report abuse