Financial Data Forecasting Using Optimized Echo State Network
The echo state network (ESN) is a dynamic neural network, which simplifies the training process in the conventional neural network. Due to its powerful non-linear computing ability, it has been applied to predict the time series. However, the parameters of the ESN need to be set experimentally, which can lead to instable performance and there is space to further improve its performance. In order to address this challenge, an improved fruit fly optimization algorithm (IFOA) is proposed in this work to optimize four key parameters of the ESN. Compared to the original fruit fly optimization algorithm (FOA), the proposed IFOA improves the optimization efficiency, where two novel particles are proposed in the fruit flies swarm, and the search process of the swarm is transformed from two-dimensional to three-dimensional space. The proposed approach is applied to financial data sets. Experimental results show that the proposed FOA-ESN and IFOA-ESN models are more effective (~50% improvement) than others, and the IFOA-ESN can obtain the best prediction accuracy.
KeywordsEcho state network Fruit fly algorithm Time series Algorithm optimization
This research is supported by the National Natural Science Foundation of China under Grant 61603104, the Guangxi Natural Science Foundation under Grants 2016GXNSFCA380017, 2015GXNSFBA139256 and 2017GXNSFAA198180, the funding of Overseas 100 Talents Program of Guangxi Higher Education, the Doctoral Research Foundation of Guangxi Normal University under Grant 2016BQ005, the Scientific Research Funds for the Returned Overseas Chinese Scholars from State Education Ministry, the Funds for Young Key Program of Education Department from Fujian Province, China (Grant No. JZ160425), and Program of Education Department of Fujian Province, China (Grant No. I201501005).
- 3.Jaeger, H.: The “echo state” approach to analysing and training recurrent neural networks. Technology GMD Technical report 148, German National Research Center Information, German (2001)Google Scholar
- 5.Steil, J.J.: Backpropagation-decorrelation: online recurrent learning with O(N) complexity. In: IEEE International Joint Conference on Neural Networks, pp. 843–848 (2004)Google Scholar
- 8.Song, Q., Zhao, X., Feng, Z., An, Y., Song, B.: Hourly electric load forecasting algorithm based on echo state neural network. In: IEEE Control and Decision Conference, pp. 3893–3897 (2011)Google Scholar
- 9.An, Y., Zhao, X., Song, Q.: Short-term traffic flow forecasting via echo state neural networks. In: IEEE Seventh International Conference on Natural Computation, pp. 844–847 (2011)Google Scholar
- 10.Embrechts, M., Alexandre, L., Linton, J.: Reservoir computing for static pattern recognition. In: European Symposium on Artificial Neural Networks, pp. 101–124 (2009)Google Scholar
- 11.Zhai, F., Lin, X., Yang, Z., Song, Y.: Notice of retraction financial time series prediction based on echo state network. In: IEEE Sixth International Conference on Natural Computation, pp. 3983–3987 (2010)Google Scholar
- 16.Wang, F.: A hybrid particle swarm algorithm with roulette selection operator. J. Southwest China Norm. Univ. 31, 93–96 (2006)Google Scholar
- 17.Shen, W., Han, Y.: Short term forecasting of Shanghai composite index based on GARCH and data mining technique. In: IEEE Circuits, Communications and System, pp. 63–66 (2010)Google Scholar
- 18.Phani, B.V., Chandra, B., Raghav, V.: Quest for efficient option pricing prediction model using machine learning techniques. In: International Joint Conference, pp. 654–657 (2011)Google Scholar