Dr. Herke van Hoof

Dr. Herke van Hoof profile

I am a researcher at the University of Amsterdam, where I focus on reinforcement learning. Currently, I am particularly interested in how reinforcement learning algorithms can make use of rich and structured prior knowledge, as captured in e.g. a knowledge graph, classical algorithms, or physics simulators.

Roles in Hybrid Intelligence: Co-Coordinator SIG Reinforcement Learning, Supervisor, former co-coordinator of the Adaptive HI research line

Hybrid Intelligence Research Objectives: Adaptive HI

Hybrid Intelligence Projects:

  • 08 UvA Hybrid approaches to learn explainable sequential behaviours
  • 17 TUD (Meta-) learning to give feedback in interactive learning
  • 2.2 VU A RL-based assistant for Patient-centric Decision Making 

Research interests: Reinforcement learning, structured data, robotics

Affiliations:

  • University of Amsterdam (UVA)

More information: website