GANStick: US stock forecasting with GAN-generated candlesticks

Man Hing Wong, Lik Hang Lee, Pan Hui

Research output: Chapter in book / Conference proceedingConference article published in proceeding or bookAcademic researchpeer-review

3 Citations (Scopus)

Abstract

Stock forecast with candlestick patterns is heavily based on template-oriented and rule-based heuristics, which requires laborious sample labelling and profound financial expertise. These methods are retrospective in nature and fail to capture premature or partial signals in candlesticks. Such rigidity limits the application of candlesticks primarily to classification tasks. Thus, we propose a novel, end-to-end deep learning model, GANStick, to address all these issues. GANStick is a conditional DCGAN-convolutional BiLSTM-based model which generates future predictive candlesticks to augment multistep time series forecasting with regression. GANStick has been empirically shown to significantly beat multiple baseline implementations, with an average error rate of 68% lower across all five timesteps on the dataset composed of 11 large-cap US stocks. GANStick is the first work in automating the workflow from candlestick pattern recognition and generation to quantifying future price volatility, with the novel generative candlestick approach using the generative adversarial network.

Original languageEnglish
Title of host publicationInternational Conference on Information Systems, ICIS 2020 - Making Digital Inclusive
Subtitle of host publicationBlending the Local and the Global
PublisherAssociation for Information Systems
Number of pages18
ISBN (Electronic)9781733632553
Publication statusPublished - Dec 2020
Externally publishedYes
Event2020 International Conference on Information Systems - Making Digital Inclusive: Blending the Local and the Global, ICIS 2020 - Virtual, Online, India
Duration: 13 Dec 202016 Dec 2020

Publication series

NameInternational Conference on Information Systems, ICIS 2020 - Making Digital Inclusive: Blending the Local and the Global

Conference

Conference2020 International Conference on Information Systems - Making Digital Inclusive: Blending the Local and the Global, ICIS 2020
Country/TerritoryIndia
CityVirtual, Online
Period13/12/2016/12/20

Keywords

  • Deep learning
  • Fintech
  • GAN
  • Machine learning
  • Stock forecasting

ASJC Scopus subject areas

  • Computer Science Applications
  • Statistics, Probability and Uncertainty
  • Library and Information Sciences
  • Applied Mathematics

Fingerprint

Dive into the research topics of 'GANStick: US stock forecasting with GAN-generated candlesticks'. Together they form a unique fingerprint.

Cite this