Open AccessProceedings Article
Fast Neighborhood Subgraph Pairwise Distance Kernel
Fabrizio Costa,Kurt De Grave +1 more
- 21 Jun 2010
- pp 255-262
TL;DR: A novel graph kernel called the Neighborhood Subgraph Pairwise Distance Kernel is introduced, which decomposes a graph into all pairs of neighborhood subgraphs of small radius at increasing distances and is shown that using a fast graph invariant the authors obtain significant speed-ups in the Gram matrix computation.
read more
Abstract: We introduce a novel graph kernel called the Neighborhood Subgraph Pairwise Distance Kernel. The kernel decomposes a graph into all pairs of neighborhood subgraphs of small radius at increasing distances. We show that using a fast graph invariant we obtain significant speed-ups in the Gram matrix computation. Finally, we test the novel kernel on a wide range of chemoinformatics tasks, from antiviral to anticarcinogenic to toxicological activity prediction, and observe competitive performance when compared against several recent graph kernel methods.
read more
Chat with Paper
AI Agents for this Paper
Find similar papers on Google Scholar, PubMed and Arxiv
Write a critical review of this paper
Analyze citations of this paper to find unaddressed research gaps
Citations
G2T-LLM: Graph-to-Tree Text Encoding for Molecule Generation with Fine-Tuned Large Language Models
Zhaoning Yu,Xiangyang Xu,Hongyang Gao +2 more
- 03 Oct 2024
TL;DR: G2T-LLM introduces graph-to-tree text encoding for molecule generation, leveraging large language models to transform complex molecular graphs into tree-structured formats, enabling intuitive interaction and generating valid chemical structures with superior generalization and adaptability.
GraphRCG: Self-conditioned Graph Generation via Bootstrapped Representations
Song Wang,Zhen Tan,Xinyu Zhao,Tianlong Chen,Huan Liu,Jundong Li +5 more
TL;DR: GraphRCG explicitly models graph distributions and uses them to guide the generation process, leading to graphs that more accurately reflect the learned distributions.
NVDiff: Graph Generation through the Diffusion of Node Vectors
TL;DR: NVDiff as mentioned in this paper proposes a score-based generative model (SGM) as a flexible prior to sample node vectors, which reduces the dimension of the diffusion process and thus improves sampling speed.
Graph Pooling via Dropping Task-Irrelevant Nodes
Cheng Zhong,Shaofeng Zhang,Feng Zhu,Rui Zhao,Xiaokang Yang,Junchi Yan +5 more
- 06 Apr 2025
TL;DR: DOTIN (Dropping Out Task-Irrelevant Nodes) is proposed to improve graph neural network scalability, achieving comparable accuracy to state-of-the-art techniques while accelerating GAT by 50% and reducing memory usage by 60% on graph-level tasks.
Extending local features with contextual information in graph kernels
TL;DR: This paper proposes a kernel based on this idea that considers trees as substructures, and where the contexts are features too, and gives an efficient algorithm for computing the kernel and shows promising results on real-world graph classification datasets.
References
Advances in kernel methods: support vector learning
Bernhard Schölkopf,Christopher John Burges,Alexander J. Smola +2 more
- 08 Feb 1999
TL;DR: Support vector machines for dynamic reconstruction of a chaotic system, Klaus-Robert Muller et al pairwise classification and support vector machines, Ulrich Kressel.
7.3K
Making large scale SVM learning practical
TL;DR: SVM light as discussed by the authors is an implementation of an SVM learner which addresses the problem of large-scale SVM training with many training examples on the shelf, which makes large scale SVM learning more practical.
5K
Convolution kernels on discrete structures
David Haussler
- 01 Jan 1999
TL;DR: A new method of constructing kernels on sets whose elements are discrete structures like strings, trees and graphs is introduced, which can be applied iteratively to build a kernel on a innnite set from kernels involving generators of the set.
1.4K
Shortest-path kernels on graphs
Karsten M. Borgwardt,Hans-Peter Kriegel +1 more
- 27 Nov 2005
TL;DR: This work proposes graph kernels based on shortest paths, which are computable in polynomial time, retain expressivity and are still positive definite, and shows significantly higher classification accuracy than walk-based kernels.
Distinguishing enzyme structures from non-enzymes without alignments.
Paul D. Dobson,Andrew J. Doig +1 more
TL;DR: It is shown that protein function can be predicted as enzymatic or not without resorting to alignments, and the method is compared to sequence-based methods that also avoid calculating alignments and predict a recently released set of unrelated proteins.
770