赵凯



摘 要:随着现在科技的快速发展,各种科技方法也随之孕育而生,例如信号处理、人工智能和模式识别等学科在各个领域都有所发展,当今的分析手段也由原先的人工分析转向智能化进行转变,但是其机构和内部的运行是较为复杂多变,现在更加强调的是处理方法和概念上改进,其中粒化计算则是强调从实际中进行处理,然后再是使用可行满意的近似解进行求解,这样来优化方法,使得方法也变得知识化,再使用一般的逻辑思维方法来求解方法。而模糊理论使用语义变量的性质,这样来处理在实际中出现的不确定性的烦恼,因此其也被应用于各大领域里。对于模糊理论的模糊时间序列的定义,对存在的模糊时间序列的模型进行分析,在这个的基础上再进一步的加入模糊时间序列的预测模型,为了更为精准的选择方法,文章先是对数据进行拟合,再选择合适的方法处理后续的工作,对于结果常常是选择MSN、平均误差和标准误差来评价一个指标。
关键词:模糊时间;计算机信息粒;平均误差;标准误差
中图分类号:TP311.13 文献标识码:A 文章编号:1001-5922(2020)10-0182-06
Abstract:With the rapid development of science and technology, a variety of scientific and technological methods are also born. For example, signal processing, artificial intelligence and pattern recognition have developed in various fields. Today's analytical methods have also changed from the original artificial analysis to the intelligence. However, their organization and internal operation are more complex and changeable. Now it is more emphasized that the processing method and concept are improved, in which granulation calculation emphasizes processing from the reality, and then using feasible and satisfactory approximate solution to solve, so as to optimize the method, make the method also become knowledge, and then use the general logical thinking method to solve the method. Fuzzy theory uses the nature of semantic variables to deal with the problems of uncertainty in practice, so it is also applied in various fields. For the definition of fuzzy time series in fuzzy theory, the existing model of fuzzy time series is analyzed, on this basis, we will further add fuzzy time series forecasting models. In order to select more accurate methods, the paper first fit the data, and then select the appropriate method to handle the subsequent work. For the results, we often choose MSN, average error and standard error to evaluate an index.
Key words:fuzzy time; computer information granule; average error; standard error
0 引言
隨着现在科技的快速发展,经济也随着增长,预测在我们实际生活中逐渐也占据着很大的地位,最为明显的一个便是在股市和人口的预测,当今股市的预测也是成为当今的研究的热点,股市中的股票指数具有随机性、时变性和非线性关系等特征,现在研究关于股票的……