Prof. Dr. Kristian Kersting

Co-Director hessian.AI

TU Darmstadt|Machine Learning

Artificial Intelligence and Machine Learning Lab (AIML) →

ORCIDOpen to collaborate

About

Kristian Kersting is Full Professor at TU Darmstadt and Co-Director of the Hessian Center for AI (hessian.AI). He is a Fellow of the European Association for Artificial Intelligence (EurAI) and recipient of the German AI Prize 2019. Before joining TU Darmstadt, he held positions at TU Dortmund, Fraunhofer IAIS, and MIT. He has published 300+ peer-reviewed papers and serves on the editorial boards of JAIR and Machine Learning.

AI Research Summary

Kristian Kersting's research bridges statistical relational learning, deep probabilistic programming, and neuro-symbolic AI. His work on lifted inference and sum-product networks has advanced efficient reasoning in high-dimensional structured domains. Recent focus areas include physics-informed machine learning and trustworthy AI.

Research Interests

Neuro-Symbolic AIPhysics-Informed Machine LearningTrustworthy AIProbabilistic CircuitsLifted InferenceSum-Product NetworksClimate AI

h-index

18,400

Citations

Patents

Active Grants

PhD Students

Spinout Signal

AI-assessed commercialization potential

7/10

Patent Activity

Industry Citations

Commercial Papers

Collaboration

Featured Paper

Physics-Informed Machine Learning for Climate Science

K. Kersting, M. Buckley, N. Lawrence

arXiv preprint, 202528 citationsPreprintCommercial relevance: 8/10

Abstract

We present a framework for incorporating physical constraints into deep learning models for climate prediction. By embedding conservation laws, symmetry principles, and thermodynamic constraints directly into the network architecture, we achieve 40% reduction in prediction error for regional climate projections while maintaining physical consistency. The approach is demonstrated on precipitation forecasting, wind energy prediction, and extreme weather event detection across European regions.

Paper Preview

Research Agent

AI-powered paper analysis

Beta

I've analyzed "Physics-Informed Machine Learning for Climate Science" by Prof. Dr. Kristian Kersting. I can help you understand the key findings, assess commercial potential, compare with competitors, or identify funding opportunities. What would you like to know?

Selected Publications (3)

Physics-Informed Machine Learning for Climate Science

K. Kersting, M. Buckley, N. Lawrence

arXiv preprint, 202528 citationsPreprint

CRS

From Statistical Relational to Neurosymbolic Artificial Intelligence

L. De Raedt, S. Dumancic, R. Manhaeve, G. Marra

Artificial Intelligence, 2024312 citations

CRS

Einsum Networks: Fast and Scalable Learning of Tractable Probabilistic Circuits

R. Peharz, S. Lang, A. Vergari, K. Kersting

ICML, 2023145 citations

CRS

CRS = Commercial Relevance Score (AI-assessed, 1-10)