基于信息分配技术的双变量模糊时间序列模型构建

Construction of Bivariate Fuzzy Time Series Model Based on Information Distribution Technology

针对目前信息分配模糊时间序列模型只能研究单变量的局限性,文章构建了一个基于信息分配技术的双变量模糊时间序列预测模型,并探讨模糊区间长度对模型预测精度的影响。以中国的互联网用户渗透率以及实际GDP为例验证模型的有效性,并选取经典马尔可夫模型作为对比模型。结果表明:模糊区间长度对信息分配模型的预测精度有影响,且模糊区间长度减小能提高预测精度。

For the limitations that fuzzy time series model of information distribution can only study univariate, this paper constructs a bivariate forecasting model of fuzzy time series based on information distribution technology, and discusses the influence of fuzzy interval length on model prediction accuracy. And then the paper takes China's Internet user penetration rate and real GDP as examples to verify the validity of the model, and also selects the classical Markov model as the contrast model. The results show that the length of fuzzy interval has an effect on the prediction accuracy of information distribution model, and the prediction accuracy can be improved by decreasing the length of fuzzy interval.