Hilbert Johan Kappen
Hilbert Johan Kappen publishes as Bert or H. J. Kappen and works on the physics of machine learning. His work covers Bayesian inference, neural networks, stochastic control and quantum learning. His path-integral control work links a class of control problems with tools from statistical physics and approximate inference. He began studying quantum learning in 2018 and collaborated on atomic-scale stochastic neural networks from 2020. His projects use physical ideas to build computational methods for AI.
- Works
- 205
- Citations
- 4,423
- h-index
- 33
Publication numbers from OpenAlex, as of 14 October 2026.
Timeline
-
As of October 2026
Bayesian learning and inference
Kappen publishes as Bert (HJ) Kappen and develops computational methods for Bayesian learning and reasoning. Exact inference can be difficult, and approximate methods from statistical physics can support learning systems.
-
As of October 2026
Path-integral control
His work connects a class of stochastic-control problems to path integrals and approximate inference. The method concerns systems such as robots that choose actions under uncertainty. Its mathematical conditions limit where the formulation applies.
-
From 2018
Quantum learning research
Since 2018, he has studied how quantum formalism can advance machine learning. His work includes quantum Boltzmann machines and learning Hamiltonians from data.
-
From 2020
Atomic-scale neural-network collaboration
He collaborated with Alex Khajetoorians’ scanning-tunneling-microscopy group from 2020. The collaboration built stochastic neural networks from switching atomic spins.
Sources and credits
Photo: ELLIS profile; creator not stated. License: Not stated by source. Source: ellis.eu. Cropped to a square and resized. Ask us to remove or replace this photo.