Model-Free Episodic Control

• DOI: 10.48550/arxiv.1606.04460
• Publication Date: 2016-01-01
• Authors: Charles Blundell; Benigno Uría; Alexander Pritzel; Yazhe Li; Avraham Ruderman; Joel Z. Leibo; Jack W. Rae; Daan Wierstra; Demis Hassabis
• Keywords: FOS: Computer and information sciences; Computer Science - Machine Learning; Statistics - Machine Learning; Quantitative Biology - Neurons and Cognition; FOS: Biological sciences; 0202 electrical engineering, electronic engineering, information engineering; Machine Learning (stat.ML); Neurons and Cognition (q-bio.NC); 02 engineering and technology; Machine Learning (cs.LG)

State of the art deep reinforcement learning algorithms take many millions interactions to attain human-level performance. Humans, on other hand, can very quickly exploit highly rewarding nuances an environment upon first discovery. In brain, such rapid is thought depend hippocampus and its capacity for episodic memory. Here we investigate whether a simple model hippocampal control learn solve difficult sequential decision-making tasks. We demonstrate that it not only attains strategy significantly faster than state-of-the-art algorithms, but also achieves higher overall reward some more challenging domains.