Communication-Efficient Distributed SVD via Local Power Iterations

Authors: Xiang Li, Shusen Wang, Kun Chen, Zhihua Zhang

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

Reproducibility Variable Result LLM Response
Research Type Experimental We conduct experiments to demonstrate the effectiveness of Local Power.
Researcher Affiliation Academia 1School of Mathematical Sciences, Peking University, China 2Department of Computer Science, Stevens Institute of Technology, USA.
Pseudocode Yes Algorithm 1 Local Power
Open Source Code No The paper does not include an explicit statement about releasing source code for the described methodology or a direct link to a code repository.
Open Datasets Yes We use 15 datasets available on the LIBSVM website.4 This page contains them all. https://www.csie.ntu. edu.tw/~cjlin/libsvmtools/datasets/.
Dataset Splits No The paper states that data samples are 'randomly shuffled and then partitioned among m nodes' but does not specify explicit train/validation/test dataset splits for reproducibility.
Hardware Specification No The paper does not provide specific details about the hardware used for running its experiments.
Software Dependencies No The paper does not provide specific software dependencies with version numbers needed to replicate the experiment.
Experiment Setup Yes All the algorithms start from the same initialization Y0. We fix the target rank to k = 5. We set m = max( n 1000 , 3) so that each node has s = 1, 000 samples, unless n is too small. For three variants of Local Power we fix p = 4 (without decaying p).