Distributed Ranking with Communications: Approximation Analysis and Applications

Authors: Hong Chen, Yingjie Wang, Yulong Wang, Feng Zheng7037-7045

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

Reproducibility Variable Result LLM Response
Research Type Experimental Theoretical and empirical assessments demonstrate the effectiveness of DLSRank-C under mild conditions.In this section, we evaluate DLSRank-C on some simulated and benchmark datasets to validate our theoretical findings.All experiments were implemented in MATLAB 2019b on an intel Core i7 with 16 GB memory.
Researcher Affiliation Academia 1College of Science, Huazhong Agricultural University, China 2College of Informatics, Huazhong Agricultural University, China 3Department of Computer Science and Engineering, Southern University of Science and Technology, China chenh@mail.hzau.edu.cn, yjaywang@126.com, wangyulong6251@gmail.com, zhengf@sustech.edu.cn
Pseudocode Yes Training flow: The training flow for DLSRank-C can be broken down into following five steps: Step 1 (Initialization): ... Step 2 (On each local machine): ... Step 3 (On global machine): ... Step 4 (On each local machine): ... Step 5 (On global machine): ...
Open Source Code No The paper does not provide any links to source code or explicitly state that the code for the described methodology is open-source or available.
Open Datasets Yes All data used here are freely available at: http://www.grouplens.org/taxonomy/term/14. The Movie Lens dataset consists of 25000095 anonymous ratings of 62423 movies made by 162541 Movie Lens users.We generate 10000 samples for training and 1000 samples for testing.
Dataset Splits No The paper mentions 10000 samples for training and 1000 for testing for simulated data, and an 8:2 ratio for training and testing on real-world data. However, it does not explicitly mention a separate 'validation' set or its split details.
Hardware Specification Yes All experiments were implemented in MATLAB 2019b on an intel Core i7 with 16 GB memory.
Software Dependencies Yes All experiments were implemented in MATLAB 2019b
Experiment Setup Yes The regularization parameter λ and bandwidth d are selected in the grids {10−2, 10−1, 1, 10, 100} and {1, 10, 102, 103}, respectively.