Efficient Generalized Conditional Gradient with Gradient Sliding for Composite Optimization

Authors: Yiu-ming Cheung, Jian Lou

IJCAI 2015 | Conference PDF | Archive PDF | Plain Text | LLM Run Details

Reproducibility Variable Result LLM Response
Research Type Experimental Experiments of CUR-like matrix factorization problem with group lasso penalty on four real-world datasets demonstrate the efficiency of the proposed method.
Researcher Affiliation Academia Yiu-ming Cheung1,2 and Jian Lou1 1Department of Computer Science, Hong Kong Baptist University, Hong Kong SAR, China 2United International College, Beijing Normal University Hong Kong Baptist University, Zhuhai, China ymc@comp.hkbu.edu.hk, jianlou@comp.hkbu.edu.hk
Pseudocode Yes For clarity, we summarize the general GCG-GS in Algorithm (1) and Algorithm (2). Algorithm (3) and Algorithm (4) show the implementation details of the Refined GCG-GS.
Open Source Code No The paper provides download links for datasets and a baseline algorithm (GCG TUM) used for comparison, but it does not provide concrete access to the source code for the proposed GCG-GS methodology.
Open Datasets Yes We utilized the following four real datasets as used in [Yu et al., 2014]: SRBCT, Brain Tumor 2, 9 Tumor and Leukemia 5, which are of sizes 83 2308, 50 10367, 60 5762, and 72 11225, respectively. 5Download from http://www.gems-system.org.
Dataset Splits No The paper specifies the datasets used but does not provide specific information about training, validation, or test dataset splits (e.g., percentages, sample counts, or predefined split citations).
Hardware Specification Yes This experiment was conducted by using MATLAB on a laptop computer of Intel Core i7 2.7GHz processor with 8 GB RAM.
Software Dependencies No The paper states that experiments were conducted using 'MATLAB' but does not specify a version number for MATLAB or any other software dependencies.
Experiment Setup Yes We set λ = 5 10 4 in our experiment. We set our inner loop estimation m to 3 for all four datasets. Other input sequences were assigned exactly as the theoretical part.