Detection and treatment of outliers in time series with practical application

	Detection and treatment of outliers in time series with practical application
Al Kut Journal of Economics Administrative Sciences
Article 1, Volume 0, Issue 17, 2018, Pages 318-338
Abstract
Abstract Getting estimators of high level of efficiencies is one of the most important stages of precise statistic analysis. Special attention therefore should be given to choose the suitable methods to get to that level of estimation, especially when time series data are contaminated (contain some observations not consistent with the rest of observations). Further, these data can not be completely subjected to the given presumptions in a way that ordinary methods of estimation would be far from reality. This study aims to conduct a study of the two methods in the detection and treatment of outliers in time series theoretically and practically , by using the style of appreciation hippocampus (Robust Estimation) which proposes method is based on Gumbel Distribution is the distribution limits of outliers , because outliers lead to inefficient model and analysis of the series is not good and the style is applied to some mistakes or some other grain mixed model (ARIMA) It has been shown clearly after the application software (SPSS, Minitab, Statistica, Easy fit, Gretl) from the results that the style appreciation of the hippocampus (Robust Estimation) is effective for the detection and treatment of outliers , and this method ensures that the model the correct data and estimates of parameters is biased.

Statistics Article View: 38 PDF Download: 11