LOOCV - Not with DIABLO?

Continuing the discussion from Integration with DIABLO for N-ingretaion with low sample size:

Could I confirm if DIABLO’s cross validation - by tune() and perf() - cannot be used with small sample sizes, or if that sentence was just an example and they can be used but with the LOOCV?

Thank you.

hi @studentScot

You can use tune()/perf() with leave one out cross-validation if your sample size is very small. If it is so small that the performance results are not good, then it is probably best not to tune the model and instead choose some set parameters.