QBMK: Quantum-based Matching Kernels for Un-attributed Graphs

Authors: Lu Bai, Lixin Cui, Ming Li, Yue Wang, Edwin Hancock

ICML 2024 | Conference PDF | Archive PDF | Plain Text | LLM Run Details

Reproducibility Variable Result LLM Response
Research Type Experimental Experimental evaluations on standard graph datasets demonstrate that the proposed QBMK kernel is able to outperform state-of-the-art graph kernels and graph deep learning approaches.
Researcher Affiliation Academia 1School of Artificial Intelligence, Beijing Normal University, Beijing 100875, China. 2School of Information, Central University of Finance and Economics, Beijing 100081, China. 3Zhejiang Institute of Optoelectronics, Jinhua 321004, China. 4Zhejiang Key Laboratory of Intelligent Education Technology and Application, Zhejiang Normal University, Jinhua 321004, China. 5Department of Computer Science, University of York, YO10 5GH York, UK.
Pseudocode No The paper describes its methods but does not provide any pseudocode or algorithm blocks.
Open Source Code No The paper mentions 'Libsvm: A library for support vector machines. Software available at http://www.csie.ntu.edu.tw/ cjlin/libsvm, 2011.' which refers to a third-party tool used, not the source code for the methodology described in this paper.
Open Datasets Yes We evaluate the classification performance of the proposed QBMK kernel on ten benchmark graph datasets extracted from bioinformatics (BIO) (Kersting et al., 2016) and computer vision (CV) (Biasotti et al., 2003; Escolano et al., 2011), respectively.
Dataset Splits Yes For each kernel, we perform the 10-fold cross-validation to calculate the classification accuracy associated with the standard C-SVM (Chang & Lin, 2011).
Hardware Specification No The paper does not specify any hardware used for running the experiments (e.g., GPU/CPU models, memory details).
Software Dependencies No The paper mentions 'Libsvm: A library for support vector machines. Software available at http://www.csie.ntu.edu.tw/ cjlin/libsvm, 2011.' but does not provide a specific version number for Libsvm or any other key software component.
Experiment Setup Yes For the proposed QBMK kernel, we set the parameter H as 10, this is because the required 10-layer expansion subgraphs rooted at each vertex for the QBMK kernel is able to cover most of the graph topological structures. For each kernel, we perform the 10-fold cross-validation to calculate the classification accuracy associated with the standard C-SVM (Chang & Lin, 2011). For each dataset, we utilize the optimal C-SVM parameters and run the experiment for 10 times, and calculate the averaged classification accuracies ( standard errors) in Table 4.