Online Graph Dictionary Learning

Authors: Cédric Vincent-Cuaz, Titouan Vayer, Rémi Flamary, Marco Corneli, Nicolas Courty

ICML 2021 | Conference PDF | Archive PDF | Plain Text | LLM Run Details

Reproducibility Variable Result LLM Response
Research Type Experimental We provide numerical evidences showing the interest of our approach for unsupervised embedding of graph datasets and for online graph subspace estimation and tracking. ... This section aims at illustrating the behavior of the approaches introduced so far for both clustering (Sections 4.1-4.2) and online subspace tracking (Section 4.3).
Researcher Affiliation Academia 1Univ.Cˆote d Azur, Inria, CNRS, LJAD, Maasai, Nice, France 2ENS de Lyon, LIP UMR 5668, Lyon, France 3Ecole Polytechnique, CMAP, UMR 7641, Palaiseau, France 4Univ.Cˆote d Azur, Center of Modeling, Simulation & Interaction, Nice, France 5Univ.Bretagne-Sud, CNRS, IRISA, Vannes, France.
Pseudocode Yes Algorithm 1 BCD for unmixing problem ... Algorithm 2 GDL: stochastic update of atoms {Cs}s [S]
Open Source Code Yes The code is available at https://github.com/cedricvincentcuaz/GDL.
Open Datasets Yes We considered well-known benchmark datasets divided into three categories: i) IMDB-B and IMDB-M (Yanardag & Vishwanathan, 2015) gather graphs without node attributes derived from social networks; ii) graphs with discrete attributes representing chemical compounds from MUTAG (Debnath et al., 1991) and cuneiform signs from PTC-MR (Krichene et al., 2015); iii) graphs with real vectors as attributes, namely BZR, COX2 (Sutherland et al., 2003) and PROTEINS, ENZYMES (Borgwardt & Kriegel, 2005).
Dataset Splits Yes A variable number of S = βk atoms is tested, where k denotes the number of classes and β {2, 4, 6, 8}, with a uniform number of atoms per class. When the order N of each atom is fixed, for GDL and GWF-f, it is set to the median order in the dataset. The atoms are initialized by randomly sampling graphs from the dataset with corresponding order. We tested 4 regularization coefficients for both methods. The embeddings w are then computed and used as input for a k-means algorithm. ... For each parameter configuration (number of atoms, number of nodes and regularization parameter) we run each experiment five times, independently, with different random initializations.
Hardware Specification Yes For instance, estimating embedding on dataset IMDB-M (see Section 4.2) over 12 atoms takes on average 44 ms per graph (on processor i9-9900K CPU 3.60GHz).
Software Dependencies No The paper mentions 'POT toolbox' and 'Adam algorithm' but does not specify version numbers for any software dependencies.
Experiment Setup Yes For the datasets with attributes involving FGW, we tested 15 values of the trade-off parameter α via a logspace search in (0, 0.5) and symmetrically (0.5, 1) and select the one minimizing our objectives. For our GDL methods as well as for GWF, a first step consists into learning the atoms. A variable number of S = βk atoms is tested, where k denotes the number of classes and β {2, 4, 6, 8}, with a uniform number of atoms per class. When the order N of each atom is fixed, for GDL and GWF-f, it is set to the median order in the dataset. The atoms are initialized by randomly sampling graphs from the dataset with corresponding order. We tested 4 regularization coefficients for both methods.