本文为布拉格捷克理工大学(作者:Oleg Ostashchuk)的硕士论文,共78页。
本文主要研究时间序列分析与预测问题。本文的目的是对现有的时间序列预测方法进行综述,包括必要的数据预处理步骤。本文选择了三种有发展前途的预测方法:ARIMA、人工神经网络和双指数平滑,主要任务是对所提供的数据进行数据分析,建立各自的预测模型,对实验结果进行了总结,并对进一步的改进进行了讨论。
Given thesis deals withthe problematic of time series analysis and forecasting. The aimof thesis is to survey an existingtime series forecasting methods, including necessary datapreprocessingsteps. There are selected threepromising forecasting methods, includingARIMA method, artificial neuralnetworks method and double exponential smoothingmethod.The main task of thethesis, is to perform data analysis of provided data and to developetheindividual forecasting models.At the end of the thesis, there are resultssummary and further improvements are dis-cussed.
-
引言
-
时间序列分析
-
预测方法
-
互联网流量的数据预测
-
主要实验
-
结论
下载地址:
https://url92.ctfile.com/f/1850492-520862489-4d23f4
(访问密码:3660)
更多精彩文章请关注公众号: