as I am about to write my diploma thesis about timeseries forecasting I came across the popular ARMA and ARIMA models for exactly this task. As I searched the forum i just found the following quote by Ingo about windowing and applying usual regression methods:: "Together with strong learners like SVM, this method often clearly outperforms classical methods like ARMA / ARIMA and delivers better and more robust results than neural networks for time series predictions." Now I would like to confirm this on my data and I wonder if any ARMA or ARIMA modelling operators are already present in the current RapidMiner distribution. If not it might be a good idea for me to implement them...
Re: Any Operators for AR(I)MA Timeseries Forecasting?
ARMA and ARIMA are now some good old friends on my ToDo list. I think they reside there since the late 2007, but I didn't mange to implement them. I would not be too unhappy if you would do and contribute your code
If you have any questions regarding the implementation, feel free to ask.