Sparse tree-based initialization for neural networks - Archive ouverte HAL Accéder directement au contenu
Pré-Publication, Document De Travail Année : 2022

Sparse tree-based initialization for neural networks

Résumé

Dedicated neural network (NN) architectures have been designed to handle specific data types (such as CNN for images or RNN for text), which ranks them among state-of-the-art methods for dealing with these data. Unfortunately, no architecture has been found for dealing with tabular data yet, for which tree ensemble methods (tree boosting, random forests) usually show the best predictive performances. In this work, we propose a new sparse initialization technique for (potentially deep) multilayer perceptrons (MLP): we first train a tree-based procedure to detect feature interactions and use the resulting information to initialize the network, which is subsequently trained via standard stochastic gradient strategies. Numerical experiments on several tabular data sets show that this new, simple and easy-to-use method is a solid concurrent, both in terms of generalization capacity and computation time, to default MLP initialization and even to existing complex deep learning solutions. In fact, this wise MLP initialization raises the resulting NN methods to the level of a valid competitor to gradient boosting when dealing with tabular data. Besides, such initializations are able to preserve the sparsity of weights introduced in the first layers of the network through training. This fact suggests that this new initializer operates an implicit regularization during the NN training, and emphasizes that the first layers act as a sparse feature extractor (as for convolutional layers in CNN).
Fichier principal
Vignette du fichier
main.pdf (2.8 Mo) Télécharger le fichier
Origine : Fichiers produits par l'(les) auteur(s)

Dates et versions

hal-03792008 , version 1 (29-09-2022)

Identifiants

Citer

Patrick Lutz, Ludovic Arnould, Claire Boyer, Erwan Scornet. Sparse tree-based initialization for neural networks. 2022. ⟨hal-03792008⟩
28 Consultations
89 Téléchargements

Altmetric

Partager

Gmail Facebook X LinkedIn More