Transfer Learning for Financial Time Series Forecasting

Time-series are widely used for representing non-stationary data such as weather information, health related data, economic and stock market indexes. Many statistical methods and traditional machine learning techniques are commonly used for forecasting time series. With the development of deep learning in artificial intelligence, many researchers have adopted new models from artificial neural networks for forecasting time series. However, poor performance of applying deep learning models in short time series hinders the accuracy in time series forecasting. In this paper, we propose a novel approach to alleviate this problem based on transfer learning. Existing work on transfer learning uses extracted features from a source dataset for prediction task in a target dataset. In this paper, we propose a new training strategy for time-series transfer learning with two source datasets that outperform existing approaches. The effectiveness of our approach is evaluated on financial time series extracted from stock markets. Experiment results show that transfer learning based on 2 data sets is superior than other base-line methods.

[1]  Everette S. Gardner,et al.  Exponential smoothing: The state of the art , 1985 .

[2]  Qun Dai,et al.  A novel transfer learning framework for time series forecasting , 2018, Knowl. Based Syst..

[3]  Yoshua Bengio,et al.  Greedy Layer-Wise Training of Deep Networks , 2006, NIPS.

[4]  E. S. Gardner.  EXPONENTIAL SMOOTHING: THE STATE OF THE ART, PART II , 2006 .

[5]  T. Mustafa,et al.  MPT64 antigen detection test improves routine diagnosis of extrapulmonary tuberculosis in a low-resource setting: A study from the tertiary care hospital in Zanzibar , 2018, PloS one.

[6]  Geoffrey E. Hinton,et al.  Reducing the Dimensionality of Data with Neural Networks , 2006, Science.

[7]  Evangelos Spiliotis,et al.  Statistical and Machine Learning forecasting methods: Concerns and ways forward , 2018, PloS one.

[8]  Taghi M. Khoshgoftaar,et al.  A survey of transfer learning , 2016, Journal of Big Data.

[9]  Yifan Gong,et al.  Cross-language knowledge transfer using multilingual deep neural network with shared hidden layers , 2013, 2013 IEEE International Conference on Acoustics, Speech and Signal Processing.

[10]  Kazuhiro Seki,et al.  Predicting Stock Market Trends by Recurrent Deep Neural Networks , 2014, PRICAI.

[11]  Thomas G. Dietterich,et al.  Transfer Learning with an Ensemble of Background Tasks , 2005, NIPS 2005.

[12]  Donald J. Berndt,et al.  Using Dynamic Time Warping to Find Patterns in Time Series , 1994, KDD Workshop.

[13]  Desheng Dash Wu,et al.  Financial Risk Forecast Using Machine Learning and Sentiment Analysis , 2015 .

[14]  Ngoc Thang Vu,et al.  Multilingual deep neural network based acoustic modeling for rapid language adaptation , 2014, 2014 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP).

[15]  Rob J Hyndman,et al.  Automatic Time Series Forecasting: The forecast Package for R , 2008 .

[16]  Yuan Yu,et al.  TensorFlow: A system for large-scale machine learning , 2016, OSDI.

[17]  Germain Forestier,et al.  Transfer learning for time series classification , 2018, 2018 IEEE International Conference on Big Data (Big Data).

[18]  Yue Zhang,et al.  Deep Learning for Event-Driven Stock Prediction , 2015, IJCAI.

[19]  Luís A. Alexandre,et al.  Transfer Learning Using Rotated Image Data to Improve Deep Neural Network Performance , 2014, ICIAR.

[20]  Yoshua Bengio,et al.  How transferable are features in deep neural networks? , 2014, NIPS.

[21]  Yoshua Bengio,et al.  Domain Adaptation for Large-Scale Sentiment Classification: A Deep Learning Approach , 2011, ICML.

[22]  Xiaolong Wang,et al.  A novel text mining approach to financial time series forecasting , 2012, Neurocomputing.

[23]  Quoc V. Le,et al.  Unsupervised Pretraining for Sequence to Sequence Learning , 2016, EMNLP.

[24]  Paras Mandal,et al.  A Hybrid Intelligent Model for Deterministic and Quantile Regression Approach for Probabilistic Wind Power Forecasting , 2014, IEEE Transactions on Power Systems.

[25]  Sancho Salcedo-Sanz,et al.  Short-term wind speed prediction in wind farms based on banks of support vector machines , 2011 .

[26]  Dong Yu,et al.  Deep Learning for Signal and Information Processing , 2013 .

[27]  Qiang Yang,et al.  A Survey on Transfer Learning , 2010, IEEE Transactions on Knowledge and Data Engineering.

[28]  Qinghua Hu,et al.  Transfer learning for short-term wind speed prediction with deep neural networks , 2016 .