REPLY TO THE COMMENTS ON "Local Overfitting Control via Leverages" in "Jacobian Conditioning Analysis for Model Validation" by I. Rivals and L. Personnaz

Yacine Oussar (1) , Gaétan Monari (1) , Gerard Dreyfus (1)

1 SIGMA - Laboratoire Signaux, Modèles et Apprentissage Statistique

"Jacobian Conditioning Analysis for Model Validation" by Rivals and Personnaz is a comment on Monari and Dreyfus (2002). In the present reply, we disprove their claims. We point to flawed reasoning in the theoretical comments, and to errors and inconsistencies in the numerical examples. Our replies are substantiated by seven counter-examples, inspired from real data, which show that (i) the comments on the accuracy of the computation of the leverages are unsupported, and that (ii) following the approach they advocate leads to discarding valid models, or validating overfitted models.

Domaines: Apprentissage [cs.LG]; Machine Learning [stat.ML]; Réseau de neurones [cs.NE]

Yacine Oussar, Gaétan Monari, Gerard Dreyfus. REPLY TO THE COMMENTS ON "Local Overfitting Control via Leverages" in "Jacobian Conditioning Analysis for Model Validation" by I. Rivals and L. Personnaz. Neural Computation, 2004, 16, pp.419-443. ⟨hal-00922198⟩