Efficient local search for L$_1$ and L$_2$ binary matrix factorization - Archive ouverte HAL Access content directly
Journal Articles Intelligent Data Analysis Year : 2016

Efficient local search for L$_1$ and L$_2$ binary matrix factorization

Abstract

Rank K Binary Matrix Factorization (BMF) approximates a binary matrix by the product of two binary matrices of lower rank, K. Several researchers have addressed this problem, focusing on either approximations of rank 1 or higher, using either the L 1 or L 2-norms for measuring the quality of the approximation. The rank 1 problem (for which the L 1 and L 2-norms are equivalent) has been shown to be related to the Integer Linear Programming (ILP) problem. We first show here that the alternating strategy with the L 2-norm, at the core of several methods used to solve BMF, can be reformulated as an Unconstrained Binary Quadratic Programming (UBQP) problem. This reformulation allows us to use local search procedures designed for UBQP in order to improve the solutions of BMF. We then introduce a new local search dedicated to the BMF problem. We show in particular that this solution is in average faster than the previously proposed ones. We then assess its behavior on several collections and methods and show that it significantly improves methods targeting the L 2-norms on all the datasets considered; for the L 1-norm, the improvement is also significant for real, structured datasets and for the BMF problem without the binary reconstruction constraint.
Fichier principal
Vignette du fichier
IDA_journal_2016.pdf (662.79 Ko) Télécharger le fichier
Origin : Files produced by the author(s)
Loading...

Dates and versions

hal-01405186 , version 1 (29-11-2016)

Identifiers

Cite

Hamid Mirisaee, Eric Gaussier, Alexandre Termier. Efficient local search for L$_1$ and L$_2$ binary matrix factorization. Intelligent Data Analysis, 2016, 20, pp.783 - 807. ⟨10.3233/IDA-160832⟩. ⟨hal-01405186⟩
244 View
369 Download

Altmetric

Share

Gmail Facebook X LinkedIn More