Dr. Manoj Kumar

Research Interests

My work focuses on data-driven modeling of infectious diseases, particularly malaria, where I investigate the impact of environmental factors such as climate and land use through physics-informed machine learning. I study the generalization capabilities of deep learning architectures, including neural ordinary differential equations (ODEs), to improve robustness in various tasks including disease forecasting. Additionally, I apply physics-informed techniques to optimize inventory management systems, enhancing resource allocation efficiency in dynamic supply chains. I am also exploring model compression via ergodic theory to balance computational efficiency and predictive accuracy in complex systems.

During my PhD, I developed advanced frameworks for age and size-structured population models, with applications in epidemiology and ecological systems. My approach integrates deterministic/stochastic partial differential equations (PDEs), and control theory with modern machine learning. By bridging classical mathematical methods with innovations like neural ODEs and physics-informed AI, I aim to advance predictive modeling of disease dynamics, operational systems, and intervention strategies.

Relevant Courses Taught

• Introduction to Data Analytics
• Engineering Analytics
• Introduction to Optimization
• Design of Algorithms