Intervention Analysis And Application Time Series Analysis Using For Arima(p,d,q)

The aim of the work is to fit time series as a collection of observations that show sequence in time. This means that there are equal periods of time between observations; each observation depends to some extent on the previous observation therefore random error effects appear with the passage of time. The error in sequence is called white noise error or (residual). The time series which we want its study, will be with regular rhythm (seconds, minutes, hours, days, weeks, months, years or thousands of years). Ambitious goal for time series analysis is to predict the future, and generally based on the assumptions of past and present characteristics. In time series analysis, use of ARIMA(p,d,q)(Integrate Autoregressive Moving Average Model) is a suitable way to explore any time series observations by three steps (identify, and estimation, diagnosis appropriate model for those time series).We used stationary time series model as a first step, when estimation parameters are stationary there will be a critical step. The maximum likelihood (MLE) is used to find the estimation parameters. ARIMA model was used because it gives the data values, significant statistical correlation, and with less random error. Ljung-BOX statistics give the p-value and chi-square distribution, and if the p-value is insignificance, it indicates that the residual is independent. When the absolute value for the roots in the (ARIMA) characteristic polynomial exceed one, then time series will have a constant mean.Intervention analysis is used to assess the impact of a special event on the time series. The main focus is to estimate the dynamic effect on the mean level of the series, but other effects are also considered. For this we consider two types of Intervention on a time series;(a) the pulse function and(b) the step function. Intervention analysis should be explained if there is a certain event that change a time series.