I have created a time series model and the mean of residual is -21.34 and the MAPE value is 10.64616. So with what value i shall go with and is there any option to fine tune the time series model?
MAPE (Mean Absolute Percentage Error) measures the error in absolute term, i.e. absolute value. this type of error may be misleading because the value of the deviation is expressed in absolute value (without sign).
RESIDUAL, on the other hand, expresses the difference between the observed and the expected or predicted. Some of the values may be higher and lower then the threshold, thus, giving positive and negative signs. This type of deviation measurement is more practical because it captures values above and below the threshold.
Paul Louangrath what changes i can perform in my time series model to bring down my mean Residual value and prediction to be more accurate.
Without knowing the model, it is impossible to tell what can be done to get a better result. However there are some things you can do about the time series.
1. You can difference your series
2. You can denoise.
3. You can remove seasonality.
etc.
- Badges
- Science topic
- Similar topics
- Statistics
- Time Series