Sie sind hier: Startseite Teaching SS2018 hrl-q1


Questions to HRL Paper 1

Q1: How does the H-DRLN architecture extend vanilla DQN and what is a Deep Skill Module?

Q2: How is knowledge efficiently retained in H-DRLN for lifelong learning?

Q3: How did H-DRLN perform in the Minecraft experiments (compared to the DDQN baseline)?