Study and application of time series forecasting based on rough set and Kernel method

Abstract

A support vector machine time series forecasting model based on rough set data preprocessing was proposed by combining rough set attribute reduction and support vector machine regression algorithm. First, remove the redundant attribute for forecasting from condition attribute by rough set method; then use the minimum condition attribute set obtained after the reduction and the corresponding initial data, reform a new training sample set which only retain the important attributes influencing the forecasting accuracy; study and train the support vector machine with the training sample obtained after reduction, and then input the reformed testing sample set according to the minimum condition attribute and corresponding initial data. The model was tested and the mapping relation was got between the condition attribute and forecasting variable. Eventually, power supply and demand were forecasted in this model. The average absolute error rates of power consumption of the whole society and yearly maximum load are respectively 14.21% and 13.23%. It shows that RS-SVM time series forecasting model has high forecasting accuracy.

Keywords

Kernel method / support vector machine / rough set / forecasting

References

[1] YuX. Y., LiongS. Y.. Forecasting of hydrologic time series with ridge regression in feature space [J]. Journal of Hydrology, 2007, 332: 290-302

[2] NikolayN. Y., MenezesL. M.. Sequential Bayesian Kernel modelling with non-Gaussian noise [J]. Neural Networks, 2008, 21: 36-47

[3] WuQ., LiuW.-y., YangY.-han.. Time series online prediction algorithm based on least squares support vector machine [J]. Journal of Central South University of Technology, 2007, 14(3): 442-446

[4] ZhangH.-l., ZouZ., LiJ., ChenX.-tao.. Flame image recognition of alumina rotary kiln by artificial neural network and support vector machine methods [J]. Journal of Central South University of Technology, 2008, 15(1): 39-43

[5] CHEN Fei, HAN Chong-zhao. Time series forecasting based on wavelet KPCA and support vector machine [C]// IEEE International Conference on Automation and Ogistics. 2007: 1487–1491.

[6] CAO L J, CHUA K S, GUAN L K. Combining KPCA with support vector machine for time series forecasting [J]. IEEE, 2003: 325–329.

[7] Jayadeva, KhemchandaniR., ChandraS.. Regularized least squares fuzzy support vector regression for time series forecasting [C]. 2006 International Joint Conference on Neural Networks, 2006, Vacouver, BC, Canada, Sheraton Vancouver Wall Central Hotel: 593-598

[8] ZhangG.-y., ZhangJ.. A novel RS-SVM dynamic prediction approach and its application [J]. Process Automation Instrumentation, 2005, 26(2): 17-20

[9] LiuQ.Rough sets and rough reasoning [M], 2005, Beijing, Science Press

[10] ZhuL., HeJ.-m., ChangSong.. Rough set integrated neural network model for forecasting stock price [J]. Chinese Journal of Management Science, 2002, 10(4): 7-12

[11] HiaJ. T., ChengC. H., ChuH. H., ChenJ. S.. Fuzzy time series model based on probabilistic approach and rough set rule induction for empirical research in stock markets [J]. Data & Knowledge Engineering, 2008, 67: 103-117

[12] XiaoZ., YeS.-j., ZhongB., SunC.-xin.. BP neural network with rough set for short term load forecasting [J]. Expert Systems with Applications, 2007, 10: 1-7

[13] HanS. W., KimJ. Y.. Rough set-based decision tree using a core attributes [J]. International Journal of Information Technology & Decision Making, 2008, 7(2): 275-290

[14] NIU Lin, ZHAO Jian-guo, DU Zhi-gang, JIN Xiao-ling. Application of time series forecasting algorithm via support vector machines to power system wide-area stability prediction [C]// Transmission and Distribution Conference and Exhibition: Asia and Pacific. IEEE/PES, 2005: 1–6.

[15] GuiW.-h., WangL.-y., YangC.-hua.. Intelligent prediction model of matte grade in copper flash smelting process [J]. Transactions of Nonferrous Metals Society of China, 2007, 17: 1075-1081

[16] CaoL. J., TayE. H.. Support vector machine with adaptive parameters in financial time series forecasting [J]. IEEE Transctions on Neural Networks, 2003, 6(14): 1506-1518

[17] HuangC.-feng.Research on power supply-demand early warning based on rough set and support vector machine [M], 2005, Beijing, North China Electric Power University Press

[18] LiY.-c., FangT.-j., YuE.-keng.. Study of support vector machines for short-term load forecasting [J]. Proceedings of the Csee, 2003, 23(6): 55-59