Generalized Shortest Path Kernel on Graphs

TL;DR

This paper introduces a generalized shortest path kernel for graph classification, demonstrating its superior performance over the original kernel through empirical results and theoretical analysis on random graph datasets.

Contribution

The paper proposes a novel generalized shortest path kernel that improves graph classification accuracy over existing kernels, supported by empirical and theoretical evidence.

Findings

Generalized shortest path kernel outperforms the original kernel in classification tasks.

Theoretical analysis explains the improved performance.

Empirical results on multiple datasets validate the approach.

Abstract

We consider the problem of classifying graphs using graph kernels. We define a new graph kernel, called the generalized shortest path kernel, based on the number and length of shortest paths between nodes. For our example classification problem, we consider the task of classifying random graphs from two well-known families, by the number of clusters they contain. We verify empirically that the generalized shortest path kernel outperforms the original shortest path kernel on a number of datasets. We give a theoretical analysis for explaining our experimental results. In particular, we estimate distributions of the expected feature vectors for the shortest path kernel and the generalized shortest path kernel, and we show some evidence explaining why our graph kernel outperforms the shortest path kernel for our graph classification problem.