Timezone: »

Learn What Not to Learn: Action Elimination with Deep Reinforcement Learning
Tom Zahavy · Matan Haroush · Nadav Merlis · Daniel J Mankowitz · Shie Mannor

Wed Dec 05 02:00 PM -- 04:00 PM (PST) @ Room 517 AB #114

Learning how to act when there are many available actions in each state is a challenging task for Reinforcement Learning (RL) agents, especially when many of the actions are redundant or irrelevant. In such cases, it is easier to learn which actions not to take. In this work, we propose the Action-Elimination Deep Q-Network (AE-DQN) architecture that combines a Deep RL algorithm with an Action Elimination Network (AEN) that eliminates sub-optimal actions. The AEN is trained to predict invalid actions, supervised by an external elimination signal provided by the environment. Simulations demonstrate a considerable speedup and added robustness over vanilla DQN in text-based games with over a thousand discrete actions.

Author Information

Tom Zahavy (Technion)

I am a research scientist @ DeepMind, where I work with Professor Satinder Baveja Singh. My research focus is on Never-Ending Reinforcement Learning. Before that, I was a Ph.D. candidate at the Technion, where Professor Shie Mannor advised me, and interned for two years at Google, where I worked with Professor Yishay Mansour and Professor Haim Kaplan.

Matan Haroush (Technion)
Nadav Merlis (Technion)
Daniel J Mankowitz (Technion)
Shie Mannor (Technion)

More from the Same Authors