استخدام طريقتي بوكس جنكيز و التنقية المكيفة لتحديد أفضل نموذج للتنبؤ بتلوث مياه نهر دجلة باستخدام السلاسل الزمنية الفئوية

Abstract EN

The categorical time series is very important because there are many important applications in the various fields of applications.

In fact, there is a clear disregard for them in the literature of the time series because of the lack of techniques to treat with these categorical time series .

Predicting its future direction also allows us to know the behavior of the phenomenon ,the objective of this study is to predict by the categorical time series of the state of pollution of the water of the Tigris River through rainwater drainage networks because the urban population exceeded these networks for the drainage of rain water only in Wasit Governorate, by using Box–Jenkins method and Adaptive Filtering method ,the parameters were estimated using the maximum likelihood method ,Which were applied to weekly data and for a full year of 51 weeks taken from environmental and private pollution laboratories in Wasit Governorate, The results of the application showed that the appropriate and efficient model for representing The categorical time series data is ARMA(4,2,1) ,The results of the prediction showed our conclusion that Adaptive Filtering method showed more accurate results than Box–Jenkins method because of predictions consistent with its Original Series