Final answer:
Exponential smoothing is an autoregression model that uses past values to make future predictions. It assigns different weights to each past observation based on their recency. While the model can be seen as infinite, in practice, a finite number of observations are used.
Step-by-step explanation:
Exponential smoothing is an autoregression model because it uses past values of the variable being forecasted to make future predictions. The model assumes that the future values of the variable depend on the current value and the errors made in the previous predictions.
Exponential smoothing assigns different weights to each past observation, with the weights decreasing exponentially as the observations become older. This means that more recent observations have a greater influence on the forecasted values than older observations.
The exponential smoothing model can be seen as an infinite autoregression model because it accounts for an infinite number of lagged values in the calculation of the forecasted values. However, in practice, only a finite number of observations are used.