Predict Stock Price with Financial News Based on Recurrent Convolutional Neural Networks

Che Yu Lee, Von Wun Soo

Research output: Chapter in Book/Report/Conference proceedingConference contributionpeer-review

42 Scopus citations

Abstract

People have been interested in making profits from financial stock market prediction. However, stock market forecast has always been a challenging problem because of its uncertainty and volatility. We take a different approach by a model called recurrent convolutional neural networks (RCN) that combines the advantages of convolutions, sequence modeling, word embedding for stock price analysis and information extraction from financial news. We then combine RCN with technical analysis indicators to predict stock price. The results show that the technical analysis model combining with RCN performs better than the technical analysis alone. Besides, the prediction error of RCN is lower than that of Long-short term memory networks.

Original languageEnglish
Title of host publicationProceedings - 2017 Conference on Technologies and Applications of Artificial Intelligence, TAAI 2017
PublisherInstitute of Electrical and Electronics Engineers Inc.
Pages160-165
Number of pages6
ISBN (Electronic)9781538642030
DOIs
StatePublished - 09 05 2018
Externally publishedYes
Event2017 Conference on Technologies and Applications of Artificial Intelligence, TAAI 2017 - Taipei, Taiwan
Duration: 01 12 201703 12 2017

Publication series

NameProceedings - 2017 Conference on Technologies and Applications of Artificial Intelligence, TAAI 2017

Conference

Conference2017 Conference on Technologies and Applications of Artificial Intelligence, TAAI 2017
Country/TerritoryTaiwan
CityTaipei
Period01/12/1703/12/17

Bibliographical note

Publisher Copyright:
© 2017 IEEE.

Keywords

  • Stock price prediction
  • convolutional neural networks
  • financial news
  • long short term memory networks
  • word to vector encoding

Fingerprint

Dive into the research topics of 'Predict Stock Price with Financial News Based on Recurrent Convolutional Neural Networks'. Together they form a unique fingerprint.

Cite this