DOI QR코드

DOI QR Code

Performance Evaluation of Price-based Input Features in Stock Price Prediction using Tensorflow

텐서플로우를 이용한 주가 예측에서 가격-기반 입력 피쳐의 예측 성능 평가

  • 송유정 (숙명여자대학교 IT공학과) ;
  • 이재원 (성신여자대학교 컴퓨터공학과) ;
  • 이종우 (숙명여자대학교 IT공학과)
  • Received : 2017.08.02
  • Accepted : 2017.09.18
  • Published : 2017.11.15

Abstract

The stock price prediction for stock markets remains an unsolved problem. Although there have been various overtures and studies to predict the price of stocks scientifically, it is impossible to predict the future precisely. However, stock price predictions have been a subject of interest in a variety of related fields such as economics, mathematics, physics, and computer science. In this paper, we will study fluctuation patterns of stock prices and predict future trends using the Deep learning. Therefore, this study presents the three deep learning models using Tensorflow, an open source framework in which each learning model accepts different input features. We expand the previous study that used simple price data. We measured the performance of three predictive models increasing the number of priced-based input features. Through this experiment, we measured the performance change of the predictive model depending on the price-based input features. Finally, we compared and analyzed the experiment result to evaluate the impact of the price-based input features in stock price prediction.

과거부터 현재까지 주식시장에 대한 주가 변동 예측은 풀리지 않는 난제이다. 주가를 과학적으로 예측하기 위해 다양한 시도 및 연구들이 있어왔지만, 아직까지 정확한 미래를 예측하는 것은 불가능하다. 하지만, 주가 예측은 경제, 수학, 물리 그리고 전산학 등 여러 관련 분야에서 오랜 관심의 대상이 되어왔다. 본 논문에서는 최근 각광 받고 있는 딥러닝(Deep-Learning)을 이용하여 주가의 변동패턴을 학습하고 미래를 예측하고자한다. 본 연구에서는 오픈소스 딥러닝 프레임워크인 텐서플로우를 이용하여 총 3가지 학습 모델을 제시하였으며, 각 학습모델은 각기 다른 입력 피쳐들을 받아들여 학습을 진행한다. 입력 피쳐는 이전 연구에서 사용한 단순 가격 데이터를 확장해 입력 피쳐 개수를 증가시켜가며 실험을 하였다. 세 가지 예측 모델의 학습 성능을 측정했으며, 이를 통해 가격-기반 입력 피쳐에 따라 달라지는 예측 모델의 성능 변화 비교 분석하여 가격-기반 입력 피쳐가 주가예측에 미치는 영향을 평가하였다.

Keywords

References

  1. J. Lee, "A Stock Trading System based on Supervised Learning of Highly Volatile Stock Price Patterns," Journal of KIISE : Computing Practices and Letters, Vol. 19, No. 1, pp. 23-29, Jan. 2013. (in Korean)
  2. S. Kim, "Data Mining Tool for Stock Investors' Decision Support," The Journal of the Korea Contents Association, Vol. 12, No. 2, pp. 472-482, Feb. 2012. (in Korean) https://doi.org/10.5392/JKCA.2012.12.02.472
  3. Y. Song and W. Lee, "A Design and Implementation of Deep Learning Model for Stock Prediction using TensorFlow," Korea Computer Congress 2017, pp. 799-801, 2017. (in Korean)
  4. H. Choi, and Y. Min, "Intelligent Information System; Introduction to deep running and major issues," Korea Information Processing Society Review, Vol. 22. No. 1, pp. 7-21, 2015.
  5. Tensorflow, [Online]. Available: https://www.tensorflow.org/
  6. L. Arnat, "Stock Price Prediction by Deep Learning," Dec. 2016.
  7. Word2vec, [Online]. Available: https://en.wikipedia.org/wiki/Word2vec
  8. A. Ryo et al., "Deep learning for stock prediction using numerical and textual information," Computer and Information Science(ICIS), 2016 IEEE/ACIS 15th International Conference on. IEEE, 2016.
  9. LSTM, [Online]. Available: https://en.wikipedia.org/wiki/Long_short-term_memory
  10. Le, Quoc, and Tomas Mikolov, "Distributed representations of sentences and documents," Proc. of the 31st International Conference on Machine Learning (ICML-14), 2014.
  11. J. Lee, "A Stock Trading System based on Moving Average Patterns and Turning Point Matrix," Journal of KIISE : Computing Practices and Letters, Vol. 18, No. 7, pp. 528-532, Jul. 2012.
  12. Bollinger Band, Available: https://en.wikipedia.org/wiki/Bollinger_Bands
  13. J. Lee, "Stock price prediction model using deep learning," 2016.
  14. Theano, [Online]. Available: http://deeplearning.net/software/theano/
  15. Torch, [Online]. Available:http://torch.ch/
  16. Caffe, [Online]. Available: http://caffe.berkeleyvision.org/
  17. Jupyter notebook, [Online]. Available: http://jupyter.org/
  18. Python, [Online]. Available: https://www.python.org/
  19. Overfitting, [Online]. Available: https://en.wikipedia.org/wiki/Overfitting
  20. Dropout, [Online]. Available: https://en.wikipedia.org/wiki/Dropout_(neural_networks)

Cited by

  1. A study on novel filtering and relationship between input-features and target-vectors in a deep learning model for stock price prediction pp.1573-7497, 2018, https://doi.org/10.1007/s10489-018-1308-x