Prediction intervals should reflect model uncertainty

Open RJT1990 opened this issue 9 years ago • 1 comments

If the model is Bayesian, the intervals should account for the uncertainty in the parameters. Bootstrapping is an option for the frequentist approach.

Jul 03 '16 20:07 RJT1990

Update: now being implemented; will be in next release. Checklist is as follows:

[ ] ARIMA models
[ ] Bayesian Neural Network models
[ ] Gaussian State Space models
[ ] GAS Models
[ ] GARCH based models
[ ] GP-NARX models
[ ] Non-Gaussian State Space Models
[ ] VAR models

Nov 03 '16 19:11 RJT1990