Single Pass Entrywise-Transformed Low Rank Approximation

Authors: Yifei Jiang, Yi Li, Yiming Sun, Jiaxin Wang, David Woodruff

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

Reproducibility Variable Result LLM Response
Research Type Experimental Additionally, we empirically validate our algorithm on a real-world data set, demonstrating significant advantages over the algorithm of Liang et al. (2020) in practice.
Researcher Affiliation Academia Yifei Jiang 1 Yi Li 2 Yiming Sun 2 Jiaxin Wang 3 David P. Woodruff 4 1Tianjin University, China 2School of Physical and Mathematical Sciences, Nanyang Technological University, Singapore 3Wuhan University of Technology, China 4Department of Computer Science, Carnegie Mellon University, USA.
Pseudocode Yes Algorithm 1 Basic heavy hitter substructure
Open Source Code No The paper does not provide concrete access to source code for the described methodology. It mentions 'All omitted proofs can be found in the full version of this paper, which is given in the supplementary material', but does not mention code.
Open Datasets Yes The data we use is based on the Wikipedia data used by Liang et al. (2020).
Dataset Splits No The paper does not provide specific dataset split information (exact percentages, sample counts, or detailed splitting methodology) for training, validation, or testing.
Hardware Specification Yes All experiments are conducted under MATLAB 2019b on a laptop with a 2.20GHz CPU and 16GB RAM.
Software Dependencies Yes All experiments are conducted under MATLAB 2019b
Experiment Setup Yes We set k = 10, m = 100 and plot the error ratios of our algorithm in Figure 1.