Distributed Deep Qlearning

# Distributed Deep Q-Learning

**arXiv ID:** 1508.04186
**Authors:** Hao Yi Ong, Kevin Chavez, Augustus Hong
**Published:** 2015-08-18T01:00:32Z
**Abstract:**
We propose a distributed deep learning model to successfully learn control policies directly from high-dimensional sensory input using reinforcement learning. The model is based on the deep Q-network, a convolutional neural network trained with a variant of Q-learning. Its input is raw pixels and its output is a value function estimating future rewards from taking an action given a system state. To distribute the deep Q-network training, we adapt the DistBelief software framework to the context of efficiently training reinforcement learning agents. As a result, the method is completely asynchronous and scales well with the number of machines. We demonstrate that the deep Q-network agent, receiving only the pixels and the game score as inputs, was able to achieve reasonable success on a simple game with minimal parameter tuning.

## Skill Description

This skill is generated from the arXiv paper: Distributed Deep Q-Learning (1508.04186).

## How to Use

[To be filled in by the user or by future automation]

## References

- [arXiv:1508.04186](http://arxiv.org/abs/1508.04186v2)