High-Confident Local Structure Guided Consensus Graph Learning For Incomplete Multi-view Clustering

Authors: Shuping Zhao, Lunke Fei, Qi Lai, Jie Wen, Jinrong Cui, Tingting Chai

IJCAI 2025 | Venue PDF | Archive PDF | Plain Text | LLM Run Details

Reproducibility Variable Result LLM Response
Research Type Experimental A number of experiments have been conducted to verify the efficacy of our approach. [...] 4 Experiments and Analysis
Researcher Affiliation Academia 1Guangdong University of Technology 2Shenzhen Institute of Advanced Technology, Chinese Academy of Sciences 3Harbin Institute of Technology, Shenzhen 4South China Agricultural University 5 Harbin Institute of Technology EMAIL, EMAIL, EMAIL, EMAIL EMAIL
Pseudocode Yes Algorithm 1 Solution of HLSCG IMC
Open Source Code No No explicit statement about providing concrete access to the source code for the methodology described in this paper was found. The GitHub links provided are for datasets used in experiments, not for the authors' own implementation.
Open Datasets Yes BBCSport1: BBCSport constitutes a document database [...] 1https://github.com/GPMVCDummy/GPMVC/tree/master/ partial MV/PVC/recreate Results/data [...] The Columbia Object Image Library2 (COIL-20) comprises a total of 1,440 images [...] 2http://www.cs.columbia.edu/CAVE/software/softlib/coil20.php [...] The Caltech-101 database encompasses 101 object categories [...] [Fei-Fei et al., 2004] [...] Referring to [Zhao et al., 2016], a subset of the BUAA-visnir face database 3 was chosen [...] 3https://github.com/hdzhao/IMG/tree/master/data.
Dataset Splits No The paper only describes how incomplete multi-view data was constructed by randomly removing 10%, 30%, and 50% instances in each view, and that algorithms were run 10 times to average results. It does not provide specific training/test/validation splits for the datasets.
Hardware Specification No All experiments conducted in this study were run on MATLAB R2020a, utilizing a hardware setup with 16.0 GB of RAM and a 3.40 GHz CPU.
Software Dependencies Yes All experiments conducted in this study were run on MATLAB R2020a
Experiment Setup Yes Initialization: Initialize Z(v) via the k-nearest neighbor graph of each view; Initialize F with the eigenvalue decomposition on the Laplacian graph of each transformed complete view; C(v) 1 = 0, C(v) 2 = 0; C(v) 3 = 0; µ = 0.1. [...] Three parameters, namely λ, β, and γ, are required to be adjusted for the proposed objective function. To identify the optimal parameter combination for each database, a series of experiments were conducted on the BUAA dataset. [...] Notably, our method exhibits minimal sensitivity to λ within the range of [10 3, 1]. Furthermore, it is evident that the highest clustering accuracy for the BUAA database is achieved with β and γ in the range of [10 5, 1] and [10 5, 1], respectively.