Manifold Structured Prediction

Authors: Alessandro Rudi, Carlo Ciliberto, GianMaria Marconi, Lorenzo Rosasco

NeurIPS 2018 | Conference PDF | Archive PDF | Plain Text | LLM Run Details

Reproducibility Variable Result LLM Response
Research Type Experimental Promising experimental results on both simulated and real data complete our study.
Researcher Affiliation Academia 1INRIA Département d informatique, École Normale Supérieure PSL Research University, Paris, France. 2Department of Electrical and Electronic Engineering, Imperial College, London, UK. 3Università degli studi di Genova & Istituto Italiano di Tecnologia, Genova, Italy. 4Massachusetts Institute of Technology, Cambridge, USA.
Pseudocode No The paper mentions that 'the algorithm is recalled in ??' (referring to a missing appendix) but does not include a structured pseudocode or algorithm block within the provided text.
Open Source Code No The paper does not provide an explicit statement or link indicating the release of open-source code for the described methodology.
Open Datasets Yes We compared the performance of the manifold structured estimator proposed in this paper with the manifold regression approach in [47] on the FVC fingerprint verification challenge dataset3. The footnote refers to http://bias.csr.unibo.it/fvc2004. Also, it mentions 'multilabel benchmark datasets described in [51]' and provides the full citation for [51].
Dataset Splits Yes We generated datasets of increasing dimension m from 5 to 50, each with 1000 points for training, 100 for validation and 100 for testing.
Hardware Specification Yes We gratefully acknowledge the support of NVIDIA Corporation for the donation of the Titan Xp GPUs and the Tesla k40 GPU used for this research.
Software Dependencies No The paper mentions algorithms and methods (e.g., 'Riemannian Gradient Descent (RGD) algorithm'), but does not provide specific software names along with their version numbers required for replication.
Experiment Setup Yes The kernel bandwidth σ was chosen and the regularization parameter λ were selected by cross-validation respectively in the ranges 0.1 to 1000 and 10 6 to 1 (logarithmically spaced).