[R-sig-ME] Cross-validation

[Haneen Farah - CITG]

Is there a package cross validation for linear mixed effect models with GroupKFold cross validation?
Thanks

[[alternative HTML version deleted]]

[Paul Buerkner]

If you are willing to go Bayesian, you can use the brms package and method
kfold.

Am Mo., 12. Nov. 2018 um 22:29 Uhr schrieb Haneen Farah - CITG <

Post by Haneen Farah - CITG
Is there a package cross validation for linear mixed effect models with
GroupKFold cross validation?
Thanks
[[alternative HTML version deleted]]
_______________________________________________
https://stat.ethz.ch/mailman/listinfo/r-sig-mixed-models

[[alternative HTML version deleted]]

[Ben Bolker]

Don't know, but I would start by going through the results of

library("sos")
findFn("grouped cross validation")

and see whether anything looks good.

For simple designs it shouldn't be *too* hard to roll something from
scratch ... but admittedly a package that had already done it would be
convenient ...

Post by Haneen Farah - CITG
Is there a package cross validation for linear mixed effect models with GroupKFold cross validation?
Thanks
[[alternative HTML version deleted]]
_______________________________________________
https://stat.ethz.ch/mailman/listinfo/r-sig-mixed-models