List of Topics:
Location Research Breakthrough Possible @S-Logix pro@slogix.in

Office Address

Social List

Distributional Reinforcement Learning With Quantile Regression - 2018

Distributional Reinforcement Learning With Quantile Regression

Research Paper on Distributional Reinforcement Learning With Quantile Regression

Research Area:  Machine Learning

Abstract:

In reinforcement learning (RL), an agent interacts with the environment by taking actions and observing the next state and reward. When sampled probabilistically, these state transitions, rewards, and actions can all induce randomness in the observed long-term return. Traditionally, reinforcement learning algorithms average over this randomness to estimate the value function. In this paper, we build on recent work advocating a distributional approach to reinforcement learning in which the distribution over returns is modeled explicitly instead of only estimating the mean. That is, we examine methods of learning the value distribution instead of the value function. We give results that close a number of gaps between the theoretical and algorithmic results given by Bellemare, Dabney, and Munos (2017). First, we extend existing results to the approximate distribution setting. Second, we present a novel distributional reinforcement learning algorithm consistent with our theoretical formulation. Finally, we evaluate this new algorithm on the Atari 2600 games, observing that it significantly outperforms many of the recent improvements on DQN, including the related distributional algorithm C51.

Keywords:  
Distributional Reinforcement Learning
Quantile Regression
Machine Learning
Deep Learning

Author(s) Name:  Will Dabney,Mark Rowland ,Marc Bellemare, RĂ©mi Munos

Journal name:  

Conferrence name:  Thirty-Second AAAI Conference on Artificial Intelligence

Publisher name:  AAAI

DOI:  10.48550/arXiv.1710.10044

Volume Information:  Vol. 32 No. 1 (2018)