Trend Analysis with Effective Covariates Based On Auto Regressive- Moving Average Time Series Residuals

Determining the pattern of a time series data is commonly established through identifying trend analysis. There is a variety of regression approaches can be chosen to perform trend analysis.All regression models are differentto the choose of which confounding factor are adjusted in the model.In view of this, when one takes into account the effective covariates in the trend analysis model,different patterns of a considered time series data setare created at each time t. This study proposes a methodology for characterizing the long term evolution of particular matterto identifyair quality analysis in the presence of radium, temperature and wind direction with correlated residuals in multiple regression models. Moreover, this is interesting in case where one performs trend analysis of the evolution of particular matters in quantity to air quality with significant effective covariates. Specifically, the considered approach provides a frame work based on the Gaussian correlated residuals where they follow a stationary Auto Regressive- Moving Average (ARMA) time series model.