Rainbow DQN — what six extensions does it combine?
hardAnswer
- (1) Double DQN, (2) Dueling architecture, (3) Prioritized replay, (4) Multi-step (n-step) returns, (5) Distributional RL (C51: model return distribution not just mean), (6) Noisy nets (parametric exploration replacing epsilon-greedy).
- Hessel et al. 2018 showed each contributes; combined SOTA on Atari for its size.
- Foundation of most subsequent value-based deep RL research.
Check yourself — multiple choice
- Random
- Double + Dueling + Prioritized replay + n-step returns + Distributional (C51) + Noisy nets — SOTA on Atari at Rainbow paper era
- Same as DQN
- Not real
Rainbow: 6 DQN extensions combined; SOTA on Atari.
#value-methods#deep-rl
Practise Reinforcement Learning
214 interview questions in this topic.