Zoltan Szabo

Zoltan Szabo is a Professor of Data Science at the Department of Statistics, London School of Economics and Political Science. His research focuses on statistical machine learning, particularly kernel methods, information theoretical estimators, and scalable computation, with applications spanning safety-critical learning, style transfer, hypothesis testing, distribution regression, econometrics, and gene analysis.

• Affiliation: Department of Statistics, LSE
• Academic Rank: Professor
• Key Expertise: Kernel Methods, Information Theoretical Estimators, Scalable Computation

His work integrates theoretical rigor with practical applications, addressing challenges in safety-critical systems and developing robust nonparametric methods. Szabo has published extensively on topics like Nyström approximation, Stein discrepancy, and random Fourier features, contributing to advancements in hypothesis testing, distribution regression, and GPU-accelerated kernel techniques.

He has served as an Area Chair for top conferences (ICML, NeurIPS, AISTATS), moderated arXiv's stat.ML, and contributed to editorial roles at JMLR and ACM Transactions on Probabilistic Machine Learning. His recent articles emphasize scalable kernel methods for high-dimensional data, with applications in climate science, finance, and neuroimaging.

Scientific Awards:

• Best Paper Award, NeurIPS 2017
• HDR (Habilitation à Diriger des Recherches) with distinction, 2019
• Programme Director of MSc Data Science, LSE

As an advisor, Szabo mentors PhD students and interns in machine learning and statistics. His work often involves interdisciplinary collaboration, including grants with institutions like the Turing Institute and European Research Council.