Practical Time Series Analysis

ARIMA

ARIMA, also known as the Box-Jenkins model, is a generalization of the ARMA model by including integrated components. The integrated components are useful when data has non-stationarity, and the integrated part of ARIMA helps in reducing the non-stationarity. The ARIMA applies differencing on time series one or more times to remove non-stationarity effect. The ARIMA(p, d, q) represent the order for AR, MA, and differencing components. The major difference between ARMA and ARIMA models is the d component, which updates the series on which forecasting model is built. The d component aims to de-trend the signal to make it stationary and ARMA model can be applied to the de-trended dataset. For different values of d, the series response ...