Regularization - Between mathematical proof and magical concept

Cet événement est passé

76 y sont allés

Algolia

55 Rue d'Amsterdam · Paris

Comment nous trouver

We meet in the lobby. Ask reception for location of group.

Image du lieu de l'événement

Détails

We meet every Friday for a peer-to-peer discussion of a pre-selected machine learning research paper. The paper and discussion is in English.

Discussion:
* What is regularization?
* When is regularization useful?
* What does regularization?
* How to select regularization algorithms?
* What are the current limits of regularization?

Videos:
* Overfitting and Regularization For Deep Learning | Two Minute Papers: https://www.youtube.com/watch?v=6aF9sJrzxaM

Papers:
GENERAL KNOWLEDGE:
* Regularization in Machine Learning: https://towardsdatascience.com/regularization-in-machine-learning-76441ddcf99a

* Concept of regularisation: https://towardsdatascience.com/concept-of-regularization-28f593cf9f8c

* REGULARIZATION: An important concept in Machine Learning: https://towardsdatascience.com/regularization-an-important-concept-in-machine-learning-5891628907ea

* How to Improve a Neural Network With Regularization: https://towardsdatascience.com/how-to-improve-a-neural-network-with-regularization-8a18ecda9fe3

* Regularization techniques for Neural Networks: https://towardsdatascience.com/regularization-techniques-for-neural-networks-e55f295f2866

MATHEMATICAL CONCEPTS:
* Regularization in Statistics: https://www.stat.berkeley.edu/~bickel/Test_BickelLi.pdf

* The learning problem and regularisation: http://www.mit.edu/~9.520/fall15/slides/class02/class02.pdf

* NYTRO: When subsampling meets early stopping: http://proceedings.mlr.press/v51/camoriano16.pdf

* Online dictionary learning for sparse coding: https://www.di.ens.fr/willow/pdfs/icml09.pdf

REGULARIZATION IN DEEP LEARNING:
* Regularisation for DL: a taxonomy: https://openreview.net/pdf?id=SkHkeixAW

* Dropout: A Simple Way to Prevent Neural Networks from Overfitting: https://www.cs.toronto.edu/~hinton/absps/JMLRdropout.pdf

* Understanding Deep Learning requires rethinking generalisation: https://arxiv.org/pdf/1611.03530.pdf