Nonlinear autoregressive exogenous model

In time series modeling, a nonlinear autoregressive exogenous model (NARX) is a nonlinear autoregressive model which has exogenous inputs. This means that the model relates the present value of the time series to both:

• past values of the same series; and
• present and past values of the driving (exogenous) series.

In addition, the model contains:

• an "error" or "residual" term

which relates to the fact that knowledge of the other terms will not enable the present value of the time series to be predicted exactly.

Such a model can be stated algebraically as

$y_{t}=F(y_{t-1},y_{t-2},y_{t-3},\ldots ,u_{t-1},u_{t-2},u_{t-3},\ldots )+\varepsilon _{t}$

Here y is the variable of interest, and u is some other variable which is associated with y. In this scheme, information about u helps predict y. Here ε is the error or residual term (sometimes called noise). For example, y may be air temperature at noon, and u may be the day of the year (day-number within year).

The function F is some nonlinear function, such as a polynomial. F can be a neural network, a wavelet network, a sigmoid network and so on. To test for non linearity in a time Series, see BDS

References

• I.J. Leontaritis and S.A. Billings. "Input-output parametric models for non-linear systems. Part I: deterministic non-linear systems". Int'l J of Control 41:303-­328, 1985.

• I.J. Leontaritis and S.A. Billings. "Input-output parametric models for non-linear systems. Part II: stochastic non-linear systems". Int'l J of Control 41:329-344, 1985.