Jes Frellsen

University of Cambridge, IT University of Copenhagen

Papers

Total Citations

H-Index

About

Jes Frellsen is a leading researcher in probabilistic machine learning, with a particular focus on advancing statistical models for circular and directional data—a domain critical to fields as diverse as robotics, bioinformatics, and the social sciences. His most influential work centers on the development of the multivariate generalised von Mises distribution, a powerful probabilistic tool that extends standard modelling techniques to handle angular variables, which are often overlooked in traditional machine learning. This contribution, detailed in his highly cited 2017 paper (with over 11 citations), provides a rigorous framework for inference and applications, enabling more accurate analysis of cyclical patterns such as wind directions, protein dihedral angles, and robot orientation. Frellsen’s work bridges a crucial gap in probabilistic modelling, offering both theoretical depth and practical utility. His research has been recognized for its impact on advancing Bayesian inference for complex, non-Euclidean data, making him a key figure in the growing field of directional statistics. Through his innovative contributions, Frellsen continues to empower researchers and practitioners to tackle real-world problems involving circular variables with greater precision and insight.