Anirbit Mukherjee

Anirbit Mukherjee is a Lecturer in Machine Learning at The University of Manchester, Department of Computer Science. His research focuses on the mathematics of neural networks and deep learning, with particular interests in machine learning and robotics. He holds a Doctor of Philosophy from Johns Hopkins University and has contributed to various research projects, including work on neural networks for solving partial differential equations and generalization bounds for physics-informed neural networks. Dr. Mukherjee is also involved in the MCAIF Centre for AI Fundamentals and accepts PhD students.

Anirbit Mukherjee is a researcher in the Department of Computer Science at The University of Manchester, focusing on Deep Learning Theory and Differential Equations. His work explores the theoretical foundations of deep neural networks, including convergence guarantees for optimization algorithms like RMSProp and ADAM, and the generalization properties of deep architectures. He also investigates the mathematical underpinnings of deep learning, such as size-independent generalization bounds and the role of overparameterization. His research extends to physics-informed neural networks (PINNs), where he examines their application in solving partial differential equations (PDEs) and their convergence properties. Additionally, he studies the dynamics of neural network training and the impact of noise on training processes.