# autoregressive model

The of order $p$, denoted AR($p$), is a random process model described by

 $y_{t}=\sum_{i=1}^{p}a_{i}y_{t-i}+c+e_{t},\qquad t=1,2,\ldots$ (1)

where $a_{i}$ are model parameters, $y_{t}$ is the model output in discrete time instant $t$. Term $c$ is an absolute term (constant) and $e_{t}$ denotes discrete white noise.

A first-order autoregression model AR(1) in the form $y_{t}=ay_{t-1}+c+e_{t}$ is one major example.

Title autoregressive model AutoregressiveModel 2013-03-22 18:33:37 2013-03-22 18:33:37 camillio (22337) camillio (22337) 5 camillio (22337) Definition msc 60G10 msc 62J05 AR