Construction of Multivariable Fuzzy Time Series Model Based on Multidimensional Information Distribution Technology
In order to enhance the forecast accuracy of multivariable fuzzy time series model for a small sample, a new multivariable fuzzy time series forecasting model was built based on the multidimensional information distribution technology. Then as an example, three variables (the time series data of total energy consumption, per capita GDP and SO2 emissions from 2001 to 2017 in China) were selected for the case analysis, which was used to verify the feasibility and to discuss the influence of the variation of fuzziness on the forecast accuracy of the model. Furthermore, a comparative analysis with the Markov model is made. The results show that (1) the suggested model can make up for the defects of small sample; (2) the predictive accuracy increases with the decrease of fuzziness; (3) the proposed model has higher forecast accuracy than the Markov model in forecasting SO2 emissions .
KeywordsMultidimensional information distribution Fuzzy approximate reasoning Multivariable fuzzy time series Emission of SO2
This research is financially supported by Program for the soft science of Shanxi Province in China (No. 2017041025-2); and Program for the Philosophy and Social Sciences Research of Higher Learning Institutions of Shanxi (PSSR) in China (No. 2017314).
- 4.Zecchin, C., Facchinetti, A., Sparacino, G., De Nicolao, G., Cobelli, C.: A new neural network approach for short-term glucose prediction using continuous glucose monitoring time-series and meal information. In: 2011 Annual International Conference of the IEEE Engineering in Medicine and Biology Society, pp. 5653–5656. IEEE Press, New York (2011)Google Scholar
- 7.Xue, Y.: A new fuzzy time series model based on fuzzy information optimization. Stat. Decis. 416(20), 12–15 (2014)Google Scholar
- 8.Shuai, Y., Song, T.L., Wang, J.P., Zhan, W.B.: Modified fuzzy time series model interval partitioning algorithm. Comput. Eng. Des. 38(2), 379–383 + 394 (2017)Google Scholar
- 10.Wu, B., Tse, S., Hsu, Y.: A new approach of bivariate fuzzy time series: with applications to the stock index forecasting. Int. J. Uncertainty Fuzziness Knowl. Based Syst. 12(11), 1793–1811 (2003)Google Scholar