Graph Invariant Kernels
نویسندگان
چکیده
We introduce a novel kernel that upgrades the Weisfeiler-Lehman and other graph kernels to effectively exploit highdimensional and continuous vertex attributes. Graphs are first decomposed into subgraphs. Vertices of the subgraphs are then compared by a kernel that combines the similarity of their labels and the similarity of their structural role, using a suitable vertex invariant. By changing this invariant we obtain a family of graph kernels which includes generalizations of Weisfeiler-Lehman, NSPDK, and propagation kernels. We demonstrate empirically that these kernels obtain state-ofthe-art results on relational data sets.
منابع مشابه
kProlog: an algebraic Prolog for kernel programming
kProlog is a simple algebraic extension of Prolog with facts and rules annotated with semiring labels. We propose kProlog as a language for learning with kernels. kProlog allows to elegantly specify systems of algebraic expressions on databases. We propose some code examples of gradually increasing complexity, we give a declarative specification of some matrix operations and an algorithm to sol...
متن کاملGraph Kernels via Functional Embedding
We propose a representation of graph as a functional object derived from the power iteration of the underlying adjacency matrix. The proposed functional representation is a graph invariant, i.e., the functional remains unchanged under any reordering of the vertices. This property eliminates the difficulty of handling exponentially many isomorphic forms. Bhattacharyya kernel constructed between ...
متن کاملA note on the bounds of Laplacian-energy-like-invariant
The Laplacian-energy-like of a simple connected graph G is defined as LEL:=LEL(G)=∑_(i=1)^n√(μ_i ), Where μ_1 (G)≥μ_2 (G)≥⋯≥μ_n (G)=0 are the Laplacian eigenvalues of the graph G. Some upper and lower bounds for LEL are presented in this note. Moreover, throughout this work, some results related to lower bound of spectral radius of graph are obtained using the term of ΔG as the num...
متن کاملA Unifying View of Explicit and Implicit Feature Maps for Structured Data: Systematic Studies of Graph Kernels
Non-linear kernel methods can be approximated by fast linear ones using suitable explicit feature maps allowing their application to large scale problems. To this end, explicit feature maps of kernels for vectorial data have been extensively studied. As many real-world data is structured, various kernels for complex data like graphs have been proposed. Indeed, many of them directly compute feat...
متن کاملOn Relation between the Kirchhoff Index and Laplacian-Energy-Like Invariant of Graphs
Let G be a simple connected graph with n ≤ 2 vertices and m edges, and let μ1 ≥ μ2 ≥...≥μn-1 >μn=0 be its Laplacian eigenvalues. The Kirchhoff index and Laplacian-energy-like invariant (LEL) of graph G are defined as Kf(G)=nΣi=1n-1<...
متن کامل