00:00:00 / 00:00:00

Apparaît dans la collection : Mathematical Methods of Modern Statistics 2 / Méthodes mathématiques en statistiques modernes 2

In this talk we consider high-dimensional classification. We discuss first high-dimensional binary classification by sparse logistic regression, propose a model/feature selection procedure based on penalized maximum likelihood with a complexity penalty on the model size and derive the non-asymptotic bounds for the resulting misclassification excess risk. Implementation of any complexity penalty-based criterion, however, requires a combinatorial search over all possible models. To find a model selection procedure computationally feasible for high-dimensional data, we consider logistic Lasso and Slope classifiers and show that they also achieve the optimal rate. We extend further the proposed approach to multiclass classification by sparse multinomial logistic regression.

This is joint work with Vadim Grinshtein and Tomer Levy.

Informations sur la vidéo

Données de citation

  • DOI 10.24350/CIRM.V.19640203
  • Citer cette vidéo Abramovich, Felix (03/06/2020). High-dimensional classification by sparse logistic regression. CIRM. Audiovisual resource. DOI: 10.24350/CIRM.V.19640203
  • URL https://dx.doi.org/10.24350/CIRM.V.19640203

Bibliographie

  • ABRAMOVICH, Felix, GRINSHTEIN, Vadim, et LEVY, Tomer. Multiclass classification by sparse multinomial logistic regression. arXiv preprint arXiv:2003.01951, 2020. - https://arxiv.org/abs/2003.01951
  • ABRAMOVICH, Felix et GRINSHTEIN, Vadim. High-dimensional classification by sparse logistic regression. IEEE Transactions on Information Theory, 2018, vol. 65, no 5, p. 3068-3079. - https://doi.org/10.1109/TIT.2018.2884963
  • ALQUIER, Pierre, COTTET, Vincent, LECUÉ, Guillaume, et al. Estimation bounds and sharp oracle inequalities of regularized procedures with Lipschitz loss functions. The Annals of Statistics, 2019, vol. 47, no 4, p. 2117-2144. - http://dx.doi.org/10.1214/18-AOS1742
  • BARTLETT, Peter L., JORDAN, Michael I., et MCAULIFFE, Jon D. Convexity, classification, and risk bounds. Journal of the American Statistical Association, 2006, vol. 101, no 473, p. 138-156. - https://www.jstor.org/stable/30047445
  • BELLEC, Pierre C., LECUÉ, Guillaume, TSYBAKOV, Alexandre B., et al. Slope meets lasso: improved oracle bounds and optimality. The Annals of Statistics, 2018, vol. 46, no 6B, p. 3603-3642. - http://dx.doi.org/10.1214/17-AOS1670
  • DANIELY, Amit, SABATO, Sivan, BEN-DAVID, Shai, et al. Multiclass learnability and the erm principle. The Journal of Machine Learning Research, 2015, vol. 16, no 1, p. 2377-2404. - http://jmlr.org/papers/volume16/daniely15a/daniely15a.pdf

Dernières questions liées sur MathOverflow

Pour poser une question, votre compte Carmin.tv doit être connecté à mathoverflow

Poser une question sur MathOverflow




Inscrivez-vous

  • Mettez des vidéos en favori
  • Ajoutez des vidéos à regarder plus tard &
    conservez votre historique de consultation
  • Commentez avec la communauté
    scientifique
  • Recevez des notifications de mise à jour
    de vos sujets favoris
Donner son avis