F. Aiolli, G. Martino, and A. Sperduti. Proceedings of the 26th International Conference on Machine Learning (ICML-09), Montreal, Canada, (2009)
Abstract
Almost all tree kernels proposed in the literature match substructures without taking into account their relative positioning with respect to one another. In this paper, we propose a novel family of kernels which explicitly focus on this type of information. Specifically, after defining a family of tree kernels based on routes between nodes, we present an efficient implementation for a member of this family. Experimental results on four different datasets show that our method is able to reach state of the art performances, obtaining in some cases performances better than computationally more demanding tree kernels.
%0 Conference Paper
%1 Aiolli:EtAl:09
%A Aiolli, Fabio
%A Martino, Giovanni Da San
%A Sperduti, Alessandro
%B Proceedings of the 26th International Conference on Machine Learning (ICML-09)
%C Montreal, Canada
%D 2009
%K 2009 icml kernels tree
%T Route Kernels for Trees
%U http://www.cs.mcgill.ca/~icml2009/papers/542.pdf
%X Almost all tree kernels proposed in the literature match substructures without taking into account their relative positioning with respect to one another. In this paper, we propose a novel family of kernels which explicitly focus on this type of information. Specifically, after defining a family of tree kernels based on routes between nodes, we present an efficient implementation for a member of this family. Experimental results on four different datasets show that our method is able to reach state of the art performances, obtaining in some cases performances better than computationally more demanding tree kernels.
@inproceedings{Aiolli:EtAl:09,
abstract = {Almost all tree kernels proposed in the literature match substructures without taking into account their relative positioning with respect to one another. In this paper, we propose a novel family of kernels which explicitly focus on this type of information. Specifically, after defining a family of tree kernels based on routes between nodes, we present an efficient implementation for a member of this family. Experimental results on four different datasets show that our method is able to reach state of the art performances, obtaining in some cases performances better than computationally more demanding tree kernels.},
added-at = {2009-06-11T15:45:14.000+0200},
address = {Montreal, Canada},
author = {Aiolli, Fabio and Martino, Giovanni Da San and Sperduti, Alessandro},
biburl = {https://www.bibsonomy.org/bibtex/27814db4799ea4cb31bae36f121187372/seandalai},
booktitle = {Proceedings of the 26th International Conference on Machine Learning (ICML-09)},
interhash = {d4247f37126ab288ed34b206fc6041ca},
intrahash = {7814db4799ea4cb31bae36f121187372},
keywords = {2009 icml kernels tree},
timestamp = {2009-06-11T15:45:14.000+0200},
title = {Route Kernels for Trees},
url = {http://www.cs.mcgill.ca/~icml2009/papers/542.pdf},
year = 2009
}