The purpose of this reppository is learning to identify an Auto-Regressive process. In order to do that, the definition and the process to identify this sort of series is presented. The next diagram shows the general process:
An AR(2) process will be simulated using the equation:
- Gather data: Simulate de AR(2) model with the above equation.
- Test stationary.
- Apply transformations until our series is stationary.
- Plot ACF: Is it slowly decaying?
- Plot the PACF:get the order of our AR model.
- Make forecasts over the test: use rolling forecast with a window lenght of the AR's order.
- Plot forecasting.
Bibliography:
- Peixeiro, M. (2022). Time Series Forecasting in Python (1st ed., Chapter 4, pp. 81-100). Manning Publications.