Why can leave-one-out yield an unstable performance estimate despite using almost all observations to train each model?
Classic MLLevel: AdvancedLeave-one-out instability
Why can leave-one-out yield an unstable performance estimate despite using almost all observations to train each model?
Hints
- Training-set size controls only part of the trade-off.
- Fold estimates are highly dependent.
Answer
This answer is part of mentoring
Advanced questions come with a full breakdown and a mock interview with a working quant researcher. The first call is free.
Source and licence: scikit-learn 1.7.2 — cross validation · BSD-3-Clause · answer: Quantoros editorial