Random extrapolation for primal-dual coordinate descent

Authors: Ahmet Alacaoglu, Olivier Fercoq, Volkan Cevher

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

Reproducibility Variable Result LLM Response
Research Type Experimental Numerical evidence demonstrates the state-of-the-art empirical performance of our method in sparse and dense settings, matching and improving the existing methods. 6. Numerical experiments
Researcher Affiliation Academia 1EPFL, Switzerland 2LTCI, T el ecom Paris, Institut Polytechnique de Paris, France.
Pseudocode Yes Algorithm 1 Primal-dual method with random extrapolation and coordinate descent (PURE-CD)
Open Source Code No The paper mentions that they implemented SPDHG and PURE-CD but does not provide any concrete access (link, explicit statement of release) to the source code for their own implementation.
Open Datasets Yes We use datasets from LIBSVM with different sparsity levels (Chang & Lin, 2011).
Dataset Splits No The paper mentions the datasets used (LIBSVM datasets) and their properties, but it does not specify any training, validation, or test splits, nor does it refer to standard predefined splits for these datasets.
Hardware Specification No The paper does not provide any specific hardware details (e.g., CPU, GPU models, or cloud instance types) used for running the experiments.
Software Dependencies No The paper mentions using a 'generic coordinate descent solver, implemented in Cython,' but it does not provide specific version numbers for Cython or any other software dependencies, libraries, or frameworks used in the experiments.
Experiment Setup Yes We select uniform sampling, pi = 1/n... We use the following step sizes, for γ < 1 σj = 1 θj maxi Ai , τi = γ maxi Ai ... We solve Lasso and ridge regression, where we let g(x) = λ x 1... and λ = 10 or λ = 10 1.