Deep Reinforcement Learning with a Natural Language Action Space (2015-11-14T00:00:00.000000Z)

TL

TL;DR

This paper introduces a novel architecture for reinforcement learning with deep neural networks designed to handle state and action spaces characterized by natural language, as found in text-based games, called a deep reinforcement relevance network (DRRN).

Abstract

This paper introduces a novel architecture for reinforcement learning with deep neural networks designed to handle state and action spaces characterized by natural language, as found in text-based games. Termed a deep reinforcement relevance network (DRRN), the architecture represents action and state spaces with separate embedding vectors, which are combined with an interaction function to approximate the Q-function in reinforcement learning. We evaluate the DRRN on two popular text games, showing superior performance over other deep Q-learning architectures. Experiments with paraphrased action descriptions show that the model is extracting meaning rather than simply memorizing strings of text.

Authors

L. Deng

8 papers

Jianfeng Gao

39 papers

Xiaodong He

16 papers

References30 items

1

WebNav: A New Large-Scale Task for Natural Language based Sequential Decision Making

2

End-to-End Goal-Driven Web Navigation

3

Continuous control with deep reinforcement learning

4

Effective Approaches to Attention-based Neural Machine Translation

5

TL;DR

Abstract

Authors

References30 items

WebNav: A New Large-Scale Task for Natural Language based Sequential Decision Making

End-to-End Goal-Driven Web Navigation

Continuous control with deep reinforcement learning

Effective Approaches to Attention-based Neural Machine Translation

Language Understanding for Text-based Games using Deep Reinforcement Learning

Skip-Thought Vectors

Human-level control through deep reinforcement learning

GloVe: Global Vectors for Word Representation

Sequence to Sequence Learning with Neural Networks

Distributed Representations of Sentences and Documents

Playing Atari with Deep Reinforcement Learning

Learning deep structured semantic models for web search using clickthrough data

Efficient Estimation of Word Representations in Vector Space

POMDP-Based Statistical Spoken Dialog Systems: A Review

ImageNet classification with deep convolutional neural networks

Context-Dependent Pre-Trained Deep Neural Networks for Large-Vocabulary Speech Recognition

Learning to Win by Reading Manuals in a Monte-Carlo Framework

Reinforcement Learning for Mapping Instructions to Actions

A unified architecture for natural language processing: deep neural networks with multitask learning

Fundamentals of Game Design

A neural probabilistic language model

Automatic learning of dialogue strategy using dialogue simulation and reinforcement learning

Reinforcement Learning for Spoken Dialogue Systems

Temporal Difference Learning and TD-Gammon

Q-learning

Deep Learning

FUNDAMENTAL TECHNOLOGIES IN MODERN SPEECH RECOGNITION Digital Object Identifier 10.1109/MSP.2012.2205597

Reinforcement learning for dialog management using least-squares Policy iteration and fast feature selection

Reinforcement Learning: An Introduction

Reinforcement Learning for Robots Using Neural Networks

Field of Study

Journal Information

Name

Venue Information

Name

Type

URL

Alternate Names