Machine learning: comment éviter le surapprentissage sur un petit dataset ?
asked 13 days ago54 viewsfr
J'ai seulement ~2000 exemples étiquetés. Le modèle mémorise l'ensemble d'entraînement. Régularisation, augmentation de données, validation croisée — par où commencer ?
🤖 AI diagnosis
AI-generated. Not an answer — the community below confirms or corrects it. Always verify before relying on it.
2 Answers
Use a healthcheck and condition: service_healthy:
db:
healthcheck:
test: ["CMD-SHELL", "pg_isready -U postgres"]
interval: 5s
retries: 5
app:
depends_on:
db:
condition: service_healthyAlso add app-level connection retries — containers restart.
Sign in to tell others whether this worked.
With ~2000 examples: start with cross-validation (5-fold), strong regularization (dropout / weight decay), and early stopping. Then data augmentation. Prefer a small model or fine-tune a pretrained one rather than training from scratch.
Sign in to tell others whether this worked.
Sign in and verify your email to post an answer.