GraKeL: A Graph Kernel Library in Python

Authors: Giannis Siglidis, Giannis Nikolentzos, Stratis Limnios, Christos Giatsidis, Konstantinos Skianis, Michalis Vazirgiannis

JMLR 2020 | Venue PDF | Archive PDF | Plain Text | LLM Run Details

Reproducibility Variable Result LLM Response
Research Type Experimental Figure 1: Running time (in seconds) for kernel computation on the ENZYMES dataset using the Gra Ke L and graphkernels libraries. ... The following example demonstrates the use of Gra Ke L for performing graph classification on a standard dataset. ... print("accuracy: %2.2f %%" %( accuracy_score (y_test , y_pred)*100)) accuracy: 84.21 %
Researcher Affiliation Academia Giannis Siglidis EMAIL LIP6, UPMC Universit e Paris 6, Sorbonne Universit es Paris, France Giannis Nikolentzos EMAIL Stratis Limnios EMAIL Christos Giatsidis EMAIL Konstantinos Skianis EMAIL LIX, Ecole Polytechnique Palaiseau, France Michalis Vazirgiannis EMAIL LIX, Ecole Polytechnique Palaiseau, France and Department of Informatics, Athens University of Economics and Business Athens, Greece
Pseudocode No The paper describes the general methods (fit, fit_transform, transform, diagonal) for the Kernel class but does not provide structured pseudocode or algorithm blocks for any specific algorithm.
Open Source Code Yes The code is BSD licensed and is available at: https://github.com/ysig/ Gra Ke L. ... Gra Ke L is a package that provides implementations of several graph kernels. The library is BSD licensed, and is publicly available on a Git Hub repository encouraging collaborative work inside the machine learning community.
Open Datasets Yes Furthermore, through its datasets submodule, Gra Ke L facilitates the application of graph kernels to several popular graph classification datasets contained in a public repository (Kersting et al., 2016). ... MUTAG = fetch_dataset("MUTAG", verbose=False)
Dataset Splits Yes G_train , G_test , y_train , y_test = train_test_split (G, y, test_size=0.1, random_state =42)
Hardware Specification No The paper discusses software libraries and benchmarks running times but does not provide specific hardware details (e.g., CPU, GPU models, memory) used for the experiments.
Software Dependencies No The paper lists several software packages as underlying technologies (NumPy, SciPy, Cython, scikit-learn, BLISS, CVXOPT) but does not provide specific version numbers for any of them.
Experiment Setup Yes The following example demonstrates the use of Gra Ke L for performing graph classification on a standard dataset. ... sp_kernel = Shortest Path () ... clf = SVC(kernel= precomputed ).fit(K_train , y_train)