Fast learning rates for plug-in classifiers
Audibert, Jean-Yves ; Tsybakov, Alexandre B.
Ann. Statist., Tome 35 (2007) no. 1, p. 608-633 / Harvested from Project Euclid
It has been recently shown that, under the margin (or low noise) assumption, there exist classifiers attaining fast rates of convergence of the excess Bayes risk, that is, rates faster than n−1/2. The work on this subject has suggested the following two conjectures: (i) the best achievable fast rate is of the order n−1, and (ii) the plug-in classifiers generally converge more slowly than the classifiers based on empirical risk minimization. We show that both conjectures are not correct. In particular, we construct plug-in classifiers that can achieve not only fast, but also super-fast rates, that is, rates faster than n−1. We establish minimax lower bounds showing that the obtained rates cannot be improved.
Publié le : 2007-04-14
Classification:  Classification,  statistical learning,  fast rates of convergence,  excess risk,  plug-in classifiers,  minimax lower bounds,  62G07,  62G08,  62H05,  68T10
@article{1183667286,
     author = {Audibert, Jean-Yves and Tsybakov, Alexandre B.},
     title = {Fast learning rates for plug-in classifiers},
     journal = {Ann. Statist.},
     volume = {35},
     number = {1},
     year = {2007},
     pages = { 608-633},
     language = {en},
     url = {http://dml.mathdoc.fr/item/1183667286}
}
Audibert, Jean-Yves; Tsybakov, Alexandre B. Fast learning rates for plug-in classifiers. Ann. Statist., Tome 35 (2007) no. 1, pp.  608-633. http://gdmltest.u-ga.fr/item/1183667286/