Tensor-Based Sketching Method for the Low-Rank Approximation of Data Streams.

Authors: Cuiyu Liu, Xiao Chuanfu, Mingshuo Ding, Chao Yang

ICLR 2023 | Conference PDF | Archive PDF | Plain Text | LLM Run Details

Reproducibility Variable Result LLM Response
Research Type Experimental A series of experiments are carried out and show that the proposed tensor-based method can be more accurate and much faster than the previous work. ... In this section, we test our algorithms and compare them to the existing data-driven algorithms for low-rank approximation of data streams. We use three datasets for comparison HSI (Imamoglu et al., 2018), Logo (Indyk et al., 2019) and MRI.
Researcher Affiliation Academia Cuiyu Liu1, Chuanfu Xiao2 3, Mingshuo Ding1, Chao Yang2 3 1Academy for Advanced Interdisciplinary Studies, Peking University, Beijing, China 2School of Mathematical Sciences, Peking University, Beijing, China 3Changsha Institute for Computing and Digital Economy, Changsha, China 2101213203@stu.pku.edu.cn, {chuanfuxiao,dingmingshuo,chao_yang}@pku.edu.cn
Pseudocode Yes Algorithm 1 The SCW algorithm (Sarlos, 2006; Clarkson & Woodruff, 2009; 2017). ... Algorithm 2 The tensor-based algorithm for low-rank approximation of the data stream D. ... Algorithm 3 The two-sided SCW algorithm. ... Algorithm 4 The two-sided tensor-based algorithm for low-rank approximation of the data stream D. ... Algorithm 5 HOOI algorithm Lathauwer et al. (2000); Kolda & Bader (2009)
Open Source Code No The paper does not provide a link to open-source code for the described methodology or explicitly state that the code is being released.
Open Datasets Yes We use three datasets for comparison HSI (Imamoglu et al., 2018), Logo (Indyk et al., 2019) and MRI. ... 1Retrieved from https://github.com/gistairc/HS-SOD. 2Retrieved from http://youtu.be/L5HQo FIa T4I. 3Retrieved from https://brainweb.bic.mni.mcgill.ca/cgi/brainweb2.
Dataset Splits No The paper mentions training and testing sets, and sample ratios, but does not explicitly state the use or size of a validation set split.
Hardware Specification Yes Experiments are run on a server equipped with an NVIDIA Tesla V100 card.
Software Dependencies No The paper mentions "Py Torch" but does not specify its version number or any other software dependencies with version numbers.
Experiment Setup Yes In all experiments, we set the rank r to 10, and the sketching size k = l = 20.