autoregressive integrated moving average (ARIMA) model

Autoregressive moving average process (ARMA) model of a differenced time series (one that has been rendered stationary by the elimination of 'drift') whose output needs to be anti-differenced to forecast the original series. ARIMA models can represent a wide range of time series data, and are used generally in computing the probability of a future value lying between any two limits. See also Box-Jenkins models.