Iterative feature selection in least square regression estimation
Pierre Alquier (2008)
Annales de l'I.H.P. Probabilités et statistiques
Similarity:
This paper presents a new algorithm to perform regression estimation, in both the inductive and transductive setting. The estimator is defined as a linear combination of functions in a given dictionary. Coefficients of the combinations are computed sequentially using projection on some simple sets. These sets are defined as confidence regions provided by a deviation (PAC) inequality on an estimator in one-dimensional models. We prove that every projection the algorithm actually improves...