Rainbow: Combining Improvements in Deep Reinforcement Learning

arxiv:1710.02298

Original source · Paper JSON

Reinforcement learning · 2017-10-06

Cites (4) · Cited by (8)

Complete cleaned Markdown · Source Markdown including bibliography

Read in parts

1. Part 1
2. Part 2
3. Part 3

Contents

• Rainbow: Combining Improvements in Deep Reinforcement Learning
• Abstract
• Introduction
• Background
• Agents and environments.
• Deep reinforcement learning and DQN.
• Extensions to DQN
• Double Q-learning.
• Prioritized replay.
• Dueling networks.
• Multi-step learning.
• Distributional RL.
• Noisy Nets.
• The Integrated Agent
• Experimental Methods
• Evaluation Methodology.
• Hyper-parameter tuning.
• Analysis
• Comparison to published baselines.
• Learning speed.
• Ablation studies.
• Discussion