I want to compare two forecasting methods using MAPE but i get errors that are above 100% for Arima, so i was wondering if this good or bad.
I also want to use the DB test, but the information on this method(DB test) seem to be scarce on the net, is the a useful source that can help me understand this?
Thanks
In choosing models, it is common practice to separate the available data into two portions, training and test data, where the training data is used to estimate any parameters of a forecasting method and the test data is used to evaluate its accuracy. Because the test data is not used in determining the forecasts, it should provide a reliable indication of how well the model is likely to forecast on new data.
Very good section presenting all the measures:
https://otexts.org/fpp2/accuracy.html
I did separate my data into training and testing sets, but i get MAP errors of +-70
we Can use forecast Skill (SS)
Forecast skills (SS) are used as typical indicators to measure the efficiency of neural networks for time series forecasting. Skill in forecasting (or skill score, forecast skill, prediction skill) is a scaled representation of forecast error that relates the forecast accuracy of a particular forecast model to some reference model
Muhammad Maladoh Bah thanks for the tip, i did try them but i was'nt planning on reporting about them
Marwan Abdul hameed Ashour are you reffering to the likes of T-test and DM-test? Could you please reference some paper on this.
- Badges
- Science topic