Raghuwansh Raj ☕️

Raghuwansh Raj

(he/him)

University of Luxembourg

Professional Summary

I’m a PhD student at the Department of Computer Science, University of Luxembourg, deep diving into the mathematical foundations of geometrical/topological deep learning. I am currently working under Professor Jun Pang.

Beyond research, I’m passionate about staying active through gym workouts, climbing, martial arts, and kickboxing. I enjoy reading novels and occasionally play Tabla, a classical Indian percussion instrument.

Download CV

Education

PhD Student in Geometrical and Topological Deep Learning.

2025-09-01
Current

FSTM, University of Luxembourg

Dual Trimester Program - Geometric Statistics

2026-09-14
2026-12-12

Hausdorff Research Institute for Mathematics (HIM), Bonn University, Germany

Double MSc in Data Science and Business Analytics.

2021-09-01
2023-03-31

Department of Data Science, CentraleSupélec & ESSEC Business School, Paris, France

Bachelor of Engineering minior in Mathematics.

2016-07-01
2020-05-31

Department of Computer Science & Engineering, Indian Institute of Technology (IIT) BHU, Varanasi, India

Interests

Geometrical & Topological Deep Learning Mathematical Foundations of GDL and TDL Non-Euclidean Deep Learning on Manifolds, Graphs, and Combinatorial Complexes Geometric Graph Neural Networks Equivariant Graph Neural Networks Expressive Power of GNNs Applications in Atomic Systems Drug Discovery Molecular Dynamics Simulations
📚 My Research
My research interests include: Geometrical & Topological Deep Learning architectures, especially on non-Euclidean domains such as manifolds, spheres, graphs, higher order data objects like combinatorial complexes. Mathematical breakthroughs in GDL/TDL through the lens of differential geometry, algebraic topology, graph theory, and group theory. Geometric graph neural networks, equivariant graph neural networks, and the expressive power of GNNs.