I have some time series datasets for prediction. In which, I got stationary after basic test by ADF() function. But I got the good result by SVR method than ARIMA. I have doubt that I can use SVR, ANN, LSTM for stationary time series data because of the nonlinear model.
Thank You
Hi,
Yes, you can use SVR, ANN, LSTM for stationary time series, you can have a look at this good book:
https://www.otexts.org/fpp
Thank You Ayman, I will go through this book. Surely, this book will help more.
Thanks Peter for valuable answer. Can you provide me some link related to it.
Hi,
I do not have any particular reference in mind. In general, stationarity means just that time series have some convenient properties and some standard econometrics tools can be applied to them. However, the underlying dynamics of stationary time series might still be very complex, and nonlinear model can be useful. It is certainly not a problem to apply nonlinear model to stationary time series.
- Badges
- Science topic
- Similar topics
- Philosophy
- Esthetics
- Interpretation