Non-parametric Models for Non-negative Functions

Authors: Ulysse Marteau-Ferey, Francis Bach, Alessandro Rudi

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

Reproducibility Variable Result LLM Response
Research Type Experimental The paper is complemented by an experimental evaluation of the model showing its effectiveness in terms of formulation, algorithmic derivation and practical results on the problems of density estimation, regression with heteroscedastic errors, and multiple quantile regression.
Researcher Affiliation Academia Ulysse Marteau-Ferey Francis Bach Alessandro Rudi INRIA École Normale Supérieure PSL Reasearch University
Pseudocode No The paper discusses algorithmic approaches (e.g., FISTA) but does not include a structured pseudocode block or a clearly labeled algorithm figure.
Open Source Code Yes The code for these experiments is available on Git Hub (https://github.com/umarteau/non_negative_model).
Open Datasets No The paper describes generating synthetic data for its experiments ('n = 50 i.i.d. points sampled from (x) = 1/2N(-1, 0.3) + 1/2N(1, 0.3)', 'data are sampled', 'a given conditional distribution P(Y|x)') but does not provide access information (link, DOI, formal citation) for a publicly available dataset.
Dataset Splits No The paper states 'Full cross-validation has been applied to each model independently, to find the best λ, λ1, λ2' but does not specify the type (e.g., k-fold) or the exact split percentages or sample counts for reproducibility.
Hardware Specification No The paper does not provide specific details regarding the hardware (e.g., GPU, CPU models, or cloud computing instances) used for running the experiments.
Software Dependencies No The paper mentions using FISTA for optimization but does not provide specific software dependencies (e.g., programming languages, libraries, or solvers) with version numbers.
Experiment Setup Yes For all methods we used (A) = λ1k Ak + λ2 F or (w) = λ kwk2. We used the Gaussian kernel k(x, x0) = exp( kx x0k2/(2σ2)) with width σ. Full cross-validation has been applied to each model independently, to find the best λ, λ1, λ2 (see Appendix E).