Distillation of Deep Reinforcement Learning Models using Fuzzy Inference Systems
Recently, significant progress has been made in the field of Deep Reinforce-
ment Learning. Many of these advances have been made using deep neural networks, which are widely regarded as black boxes. In this work, we use policy distillation to distill the learned policy from a deep Q-network to a fuzzy controller, which has less black-box characteristics than a neural network.
Arne Gevaert graduated from Ghent University in July, and started his PhD in October of 2019. His work focuses on interpretability in Machine Learning, more specifically extracting interpretable knowledge from Deep Learning models.