Stochastic Optimization for Regularized Wasserstein Estimators

Authors: Marin Ballu, Quentin Berthet, Francis Bach

ICML 2020 | Conference PDF | Archive PDF | Plain Text | LLM Run Details

Reproducibility Variable Result LLM Response
Research Type Experimental We provide theoretical guarantees and illustrate the performance of our algorithm with experiments on synthetic data. We demonstrate the performance of the algorithm on simulated experiments.
Researcher Affiliation Collaboration 1Department of Pure Mathematics and Mathematical Statistics, University of Cambridge, United Kingdom 2Google Research, Brain Team, Paris, France 3Inria, ENS, PSL Research University, Paris, France.
Pseudocode Yes Algorithm 1 SGD for Wasserstein estimator
Open Source Code No The paper does not contain a statement or link indicating the release of source code for the described methodology.
Open Datasets No The paper uses synthetic data, stating, "We generate X and Y randomly by drawing two sets of independent Gaussian vectors of respective sizes I and J." However, it does not provide access information (e.g., a link or citation) for this data or the generation process.
Dataset Splits No The paper describes simulations with synthetic data but does not specify explicit training, validation, or test splits for this data.
Hardware Specification No The paper does not contain any specific details about the hardware (e.g., GPU/CPU models, memory) used to run the experiments.
Software Dependencies No The paper does not list specific software dependencies (e.g., libraries, frameworks, or solvers) along with their version numbers.
Experiment Setup Yes Here ε = 0.01, η = 0.02, with the same problem is the same as in the experiments on the regularization term shown in Figure 1.