What is a ARIMA model ?
Autoregressive Integrated Moving Average (ARIMA).
An Autoregressive integrated moving average model is more of a general form of an autoregressive moving average (ARMA) model. Each of those models is fitted to time series data either to better perceive the data or to predict future points within the series. ARIMA models are applied in some cases wherever data show proof of non-stationary within the sense of mean, an Associate in Nursing initial differencing step is applied one or many times to eliminate the non-stationary of the mean function (i.e., the trend). once the seasonality shows in an exceedingly time series, the seasonal-differencing may be applied to eliminate the seasonal element. Since the ARMA model, in keeping with Wold's decomposition theorem, is theoretically enough to explain a wide-sense stationary time series, we are impelled to create stationary a non-stationary statistic.
The AR a part of ARIMA indicates that the evolving variable of interest is regressed on its own lagged values. The MA half indicates that the regression error is really a linear combination of error terms whose values occurred contemporaneously and at varied times within the past. The I (for "integrated") indicates that the data values are replaced with the difference between their values and therefore the previous values. The aim of each of these variables is to create a model that fits the data potentially.