Similarity Learning for High-Dimensional Sparse Data - Inria - Institut national de recherche en sciences et technologies du numérique Access content directly
Conference Papers Year : 2015

Similarity Learning for High-Dimensional Sparse Data


A good measure of similarity between data points is crucial to many tasks in machine learning. Similarity and metric learning methods learn such measures automatically from data, but they do not scale well respect to the dimensionality of the data. In this paper, we propose a method that can learn efficiently similarity measure from high-dimensional sparse data. The core idea is to parameterize the similarity measure as a convex combination of rank-one matrices with specific sparsity structures. The parameters are then optimized with an approximate Frank-Wolfe procedure to maximally satisfy relative similarity constraints on the training data. Our algorithm greedily incorporates one pair of features at a time into the similarity measure, providing an efficient way to control the number of active features and thus reduce overfitting. It enjoys very appealing convergence guarantees and its time and memory complexity depends on the spar-sity of the data instead of the dimension of the feature space. Our experiments on real-world high-dimensional datasets demonstrate its potential for classification, dimensionality reduction and data exploration.
Fichier principal
Vignette du fichier
aistats15.pdf (670.7 Ko) Télécharger le fichier
Origin : Files produced by the author(s)

Dates and versions

hal-01430854 , version 1 (10-01-2017)


  • HAL Id : hal-01430854 , version 1


Kuan Liu, Aurélien Bellet, Fei Sha. Similarity Learning for High-Dimensional Sparse Data. International Conference on Artificial Intelligence and Statistics (AISTATS 2015), May 2015, San Diego, United States. ⟨hal-01430854⟩
130 View
43 Download


Gmail Facebook X LinkedIn More