A powerful and widely used time series model that captures the dependence among observations by incorporating three components: autoregression, differencing, and moving average. ARIMA models are used for forecasting and identifying trends, cycles, and outliers in time series data.