Graph kernels based on linear patterns: Theoretical and experimental comparisons

Jia, Linlin; Gaüzère, Benoît; Honeiné, Paul

doi:10.1016/j.eswa.2021.116095

Cited by 13 publications

(13 citation statements)

References 29 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…For instance, they produce good performance, but it is difficult to prove the formal requirement. In structural pattern recognition, many of the graph kernels [13,16,27] are PSD at best. The need of indefinite kernel methods is thus a real issue.…”

Section: Need Of Indefinite Kernel Methodsmentioning

confidence: 99%

“…Despite the diversity in the design of graph kernels, their use for classification is rather monotonous, namely by using support vector machines (SVM). This is also reflected in the recent survey papers for graph kernels: "The criteria used for prediction are SVM for classification" [13]; "We performed classification experiments using the C-SVM implementation LIBSVM" [16]; "In the case of graph kernels, to perform graph classification, we employed a Support Vector Machine (SVM) classifier and in particular, the LIB-SVM implementation" [27]. This is not a surprise due to the dominance of SVM in machine learning in general.…”

Section: Introductionmentioning

confidence: 91%

“…The terminology used in our work is due to its clarity and helps to avoid potential confusion. Dozens of graph kernels have been developed [13,16,27], which focus on specific structural properties of graphs. The key design ideas behind these graph kernels differ, e.g.…”

Section: Graph Kernelsmentioning

confidence: 99%

“…Graph kernels provide a way to compute similarities between graphs [13,16,27]. As such they can be used for graph classification, in particular with Support Vector Machine (SVM) but not only.…”

Section: Introductionmentioning

confidence: 99%

See 3 more Smart Citations

Interpolation Kernel Machine and Indefinite Kernel Methods for Graph Classification

Zhang

Liu

Jiang

2022

Pattern Recognition and Artificial Intelligence

View full text Add to dashboard Cite

Graph kernels have been studied for a long time and applied among others for graph classification. In this paper we bring two novel aspects into the graph processing community. Currently, the backbone for kernel-based classification is solely the support vector machine. We introduce the interpolation kernel machine for this purpose. In addition, for both support vector machine and interpolation kernel machine, many kernels used in practice do not satisfy the formal requirements (e.g. positive definiteness). We thus introduce extensions of the standard version to indefinite kernel methods. We argue and experimentally demonstrate why these two aspects should be considered for graph classification. One of our conclusions will be that the interpolation kernel machine is a good alternative of support vector machine. Consequently, we will propose an extended experimental protocol. With this work we contribute to increasing the methodological plurality in the graph processing community.

show abstract

Section: Need Of Indefinite Kernel Methodsmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 91%

Section: Graph Kernelsmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

See 2 more Smart Citations

Interpolation Kernel Machine and Indefinite Kernel Methods for Graph Classification

Zhang

Liu

Jiang

2022

Pattern Recognition and Artificial Intelligence

View full text Add to dashboard Cite

show abstract

“…G RAPH data processing using neural networks has been broadly attracting more and more research interests recently. Graph convolutional networks (GCNs) [1], [2], [3], [4], [5], [6], [7], [8], [9], [10], [11], [12], [13], [14], [15], [16], [17], [18], [19], [20], [21], [22], [23], [24], [25], [26], [27], [28], [29], [30], [31], [32], [33], [34], [35], [36], [37], [38], [39] are a family of graphbased neural networks that extend convolutional neural networks (CNNs) to extract local features in general graphs with irregular input structures. The irregularity of a graph, including the orderless nodes and connections, however, makes the GCNs difficult to design as well as training from local patterns.…”

Section: Introductionmentioning

confidence: 99%

Revisiting 2D Convolutional Neural Networks for Graph-Based Applications

Lyu

Huang

Zhang

2023

IEEE Trans. Pattern Anal. Mach. Intell.

View full text Add to dashboard Cite

Graph convolutional networks (GCNs) are widely used in graph-based applications such as graph classification and segmentation. However, current GCNs have limitations on implementation such as network architectures due to their irregular inputs. In contrast, convolutional neural networks (CNNs) are capable of extracting rich features from large-scale input data, but they do not support general graph inputs. To bridge the gap between GCNs and CNNs, in this paper we study the problem of how to effectively and efficiently map general graphs to 2D grids that CNNs can be directly applied to, while preserving graph topology as much as possible. We therefore propose two novel graph-to-grid mapping schemes, namely, graph-preserving grid layout (GPGL) and its extension Hierarchical GPGL (H-GPGL) for computational efficiency. We formulate the GPGL problem as integer programming and further propose an approximate yet efficient solver based on a penalized Kamada-Kawai method, a well-known optimization algorithm in 2D graph drawing. We propose a novel vertex separation penalty that encourages graph vertices to lay on the grid without any overlap. Along with this image representation, even extra 2D maxpooling layers contribute to the PointNet, a widely applied point-based neural network. We demonstrate the empirical success of GPGL on general graph classification with small graphs and H-GPGL on 3D point cloud segmentation with large graphs, based on 2D CNNs including VGG16, ResNet50 and multi-scale maxout (MSM) CNN.

show abstract

Predicting redox potentials by graph‐based machine learning methods

Jia,

Brémond,

Zaida

et al. 2024

J Comput Chem

View full text Add to dashboard Cite

The evaluation of oxidation and reduction potentials is a pivotal task in various chemical fields. However, their accurate prediction by theoretical computations, which is a complementary task and sometimes the only alternative to experimental measurement, may be often resource‐intensive and time‐consuming. This paper addresses this challenge through the application of machine learning techniques, with a particular focus on graph‐based methods (such as graph edit distances, graph kernels, and graph neural networks) that are reviewed to enlighten their deep links with theoretical chemistry. To this aim, we establish the ORedOx159 database, a comprehensive, homogeneous (with reference values stemming from density functional theory calculations), and reliable resource containing 318 one‐electron reduction and oxidation reactions and featuring 159 large organic compounds. Subsequently, we provide an instructive overview of the good practice in machine learning and of commonly utilized machine learning models. We then assess their predictive performances on the ORedOx159 dataset through extensive analyses. Our simulations using descriptors that are computed in an almost instantaneous way result in a notable improvement in prediction accuracy, with mean absolute error (MAE) values equal to 5.6 kcal mol for reduction and 7.2 kcal mol for oxidation potentials, which paves a way toward efficient in silico design of new electrochemical systems.

show abstract

Graph kernels based on linear patterns: Theoretical and experimental comparisons

Cited by 13 publications

References 29 publications

Interpolation Kernel Machine and Indefinite Kernel Methods for Graph Classification

Interpolation Kernel Machine and Indefinite Kernel Methods for Graph Classification

Revisiting 2D Convolutional Neural Networks for Graph-Based Applications

Predicting redox potentials by graph‐based machine learning methods

Contact Info

Product

Resources

About