Exponential Smoothing Models

A class of time series models where the assigned weight for each previous period decreases exponentially as we go back. In other words, in these models more weight is given to recent observations and less to others. This contrasts with simple moving average models where each observation is weighted equally.