Correlated Variational Auto-Encoders

Authors: Da Tang, Dawen Liang, Tony Jebara, Nicholas Ruozzi

ICML 2019 | Conference PDF | Archive PDF | Plain Text | LLM Run Details

Reproducibility Variable Result LLM Response
Research Type Experimental Experimental results on matching and link prediction on public benchmark rating datasets and spectral clustering on a synthetic dataset show the effectiveness of the proposed method over baseline algorithms.
Researcher Affiliation Collaboration 1Department of Computer Science, Columbia University, New York, New York, USA 2Netflix Inc., Los Gatos, California, USA 3Erik Jonsson School of Engineering & Computer Science, University of Texas at Dallas, Richardson, Texas, USA.
Pseudocode Yes Algorithm 1 Computing all weights w MAS G,e
Open Source Code Yes Code is available at https://github.com/datang1992/CorrelatedVAEs.
Open Datasets Yes We use the Movie Lens 20M dataset (Harper & Konstan, 2016). (...) In this experiment, we use the Epinions dataset (Massa & Avesani, 2007)
Dataset Splits Yes For all the experiments, we did a stochastic train/test split over users with a 90/10 ratio. (...) To split the train/test dataset for the link prediction task, for each vertex vi V , we hold out max 1, 1 20 degree(vi) edges on vi as the testing edge set Etest, and put all edges that are not selected into the training edge set Etrain.
Hardware Specification No The paper does not provide specific hardware details (exact GPU/CPU models, processor types with speeds, memory amounts, or detailed computer specifications) used for running its experiments. It does not mention any specific hardware used for training or evaluation.
Software Dependencies No The paper does not provide specific ancillary software details, such as library or solver names with version numbers (e.g., Python 3.8, PyTorch 1.9, CPLEX 12.4).
Experiment Setup Yes For CVAEcorr, we set the negative sampling regularization parameter γ = 1. (...) Here we apply negative sampling regularization parameter γ = 0.1 for CVAEcorr. (...) Here we apply a large regularization value of γ = 1000 for the CVAEcorr