Interpretable Stein Goodness-of-fit Tests on Riemannian Manifold

Authors: Wenkai Xu, Takeru Matsuda

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

Reproducibility Variable Result LLM Response
Research Type Experimental Simulation results and real data applications show validity and usefulness of the proposed methods.
Researcher Affiliation Academia 1Gatsby Computational Neuroscience Unit, London, United Kingdom 2RIKEN Center for Brain Science, Tokyo, Japan.
Pseudocode Yes Algorithm 1 m KSD test via wild-bootstrap
Open Source Code No The paper does not provide any explicit statements or links indicating that the source code for the described methodology is publicly available.
Open Datasets Yes We use the vectorcardiogram data studied by Jupp et al. (2008). Data available on Japan Meteorological Agency website https://www.data.jma.go.jp/obd/stats/etrn/.
Dataset Splits No The paper specifies parameters for bootstrap sampling and significance levels, but it does not explicitly provide details about training, validation, or test dataset splits (e.g., percentages or counts).
Hardware Specification No The paper does not provide any specific details about the hardware used to conduct the experiments, such as CPU or GPU models, memory, or cloud computing instances.
Software Dependencies No The paper mentions statistical concepts and methods (e.g., Euler angle, noise contrastive estimation) and refers to existing literature, but it does not list any specific software or library names with version numbers that were used for implementation or experiments.
Experiment Setup Yes The bootstrap sample size is set to B = 1000. The significance level is set to α = 0.01. For the m KSD(0) test (MMD two-sample test), the number of samples from the null is set to be equal to the sample size n. We used the kernel k(X, Y ) = exp(γ tr(X Y )), where the parameter γ was chosen by optimizing the approximate test power.