A new tree kernel based on SOM-SD

Fabio Aiolli, Giovanni Martino, Alessandro Sperduti

Research output: Chapter in Book/Report/Conference proceedingConference contribution

3 Citations (Scopus)

Abstract

Many different paradigms have been studied in the past to treat tree structured data, including kernel and neural based approaches. However, both types of methods have their own drawbacks. Kernels typically can only cope with discrete labels and tend to be sparse. On the other side, SOM-SD, an extension of the SOM for structured data, is unsupervised and Markovian, i.e. the representation of a subtree does not consider where the subtree appears in a tree. In this paper, we present a hybrid approach which tries to overcome these problems. In particular, we propose a new kernel based on SOM-SD which adds information about the relative position of subtrees (the route) to the activation of the nodes in such a way to discriminate even those subtrees originally encoded by the same prototypes. Experiments have been performed against two well known benchmark datasets with promising results.

Original languageEnglish
Title of host publicationLecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Pages49-58
Number of pages10
Volume6353 LNCS
EditionPART 2
DOIs
Publication statusPublished - 2010
Externally publishedYes
Event20th International Conference on Artificial Neural Networks, ICANN 2010 - Thessaloniki
Duration: 15 Sep 201018 Sep 2010

Publication series

NameLecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
NumberPART 2
Volume6353 LNCS
ISSN (Print)03029743
ISSN (Electronic)16113349

Other

Other20th International Conference on Artificial Neural Networks, ICANN 2010
CityThessaloniki
Period15/9/1018/9/10

Fingerprint

Labels
Chemical activation
kernel
Experiments
Hybrid Approach
Activation
Paradigm
Prototype
Tend
Benchmark
Vertex of a graph
Experiment

ASJC Scopus subject areas

  • Computer Science(all)
  • Theoretical Computer Science

Cite this

Aiolli, F., Martino, G., & Sperduti, A. (2010). A new tree kernel based on SOM-SD. In Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics) (PART 2 ed., Vol. 6353 LNCS, pp. 49-58). (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 6353 LNCS, No. PART 2). https://doi.org/10.1007/978-3-642-15822-3_6

A new tree kernel based on SOM-SD. / Aiolli, Fabio; Martino, Giovanni; Sperduti, Alessandro.

Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics). Vol. 6353 LNCS PART 2. ed. 2010. p. 49-58 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 6353 LNCS, No. PART 2).

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Aiolli, F, Martino, G & Sperduti, A 2010, A new tree kernel based on SOM-SD. in Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics). PART 2 edn, vol. 6353 LNCS, Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), no. PART 2, vol. 6353 LNCS, pp. 49-58, 20th International Conference on Artificial Neural Networks, ICANN 2010, Thessaloniki, 15/9/10. https://doi.org/10.1007/978-3-642-15822-3_6
Aiolli F, Martino G, Sperduti A. A new tree kernel based on SOM-SD. In Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics). PART 2 ed. Vol. 6353 LNCS. 2010. p. 49-58. (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); PART 2). https://doi.org/10.1007/978-3-642-15822-3_6
Aiolli, Fabio ; Martino, Giovanni ; Sperduti, Alessandro. / A new tree kernel based on SOM-SD. Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics). Vol. 6353 LNCS PART 2. ed. 2010. pp. 49-58 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); PART 2).
@inproceedings{263a96f15bfb4144bbd96987d65e37c3,
title = "A new tree kernel based on SOM-SD",
abstract = "Many different paradigms have been studied in the past to treat tree structured data, including kernel and neural based approaches. However, both types of methods have their own drawbacks. Kernels typically can only cope with discrete labels and tend to be sparse. On the other side, SOM-SD, an extension of the SOM for structured data, is unsupervised and Markovian, i.e. the representation of a subtree does not consider where the subtree appears in a tree. In this paper, we present a hybrid approach which tries to overcome these problems. In particular, we propose a new kernel based on SOM-SD which adds information about the relative position of subtrees (the route) to the activation of the nodes in such a way to discriminate even those subtrees originally encoded by the same prototypes. Experiments have been performed against two well known benchmark datasets with promising results.",
author = "Fabio Aiolli and Giovanni Martino and Alessandro Sperduti",
year = "2010",
doi = "10.1007/978-3-642-15822-3_6",
language = "English",
isbn = "3642158218",
volume = "6353 LNCS",
series = "Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)",
number = "PART 2",
pages = "49--58",
booktitle = "Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)",
edition = "PART 2",

}

TY - GEN

T1 - A new tree kernel based on SOM-SD

AU - Aiolli, Fabio

AU - Martino, Giovanni

AU - Sperduti, Alessandro

PY - 2010

Y1 - 2010

N2 - Many different paradigms have been studied in the past to treat tree structured data, including kernel and neural based approaches. However, both types of methods have their own drawbacks. Kernels typically can only cope with discrete labels and tend to be sparse. On the other side, SOM-SD, an extension of the SOM for structured data, is unsupervised and Markovian, i.e. the representation of a subtree does not consider where the subtree appears in a tree. In this paper, we present a hybrid approach which tries to overcome these problems. In particular, we propose a new kernel based on SOM-SD which adds information about the relative position of subtrees (the route) to the activation of the nodes in such a way to discriminate even those subtrees originally encoded by the same prototypes. Experiments have been performed against two well known benchmark datasets with promising results.

AB - Many different paradigms have been studied in the past to treat tree structured data, including kernel and neural based approaches. However, both types of methods have their own drawbacks. Kernels typically can only cope with discrete labels and tend to be sparse. On the other side, SOM-SD, an extension of the SOM for structured data, is unsupervised and Markovian, i.e. the representation of a subtree does not consider where the subtree appears in a tree. In this paper, we present a hybrid approach which tries to overcome these problems. In particular, we propose a new kernel based on SOM-SD which adds information about the relative position of subtrees (the route) to the activation of the nodes in such a way to discriminate even those subtrees originally encoded by the same prototypes. Experiments have been performed against two well known benchmark datasets with promising results.

UR - http://www.scopus.com/inward/record.url?scp=78049397680&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=78049397680&partnerID=8YFLogxK

U2 - 10.1007/978-3-642-15822-3_6

DO - 10.1007/978-3-642-15822-3_6

M3 - Conference contribution

SN - 3642158218

SN - 9783642158216

VL - 6353 LNCS

T3 - Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)

SP - 49

EP - 58

BT - Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)

ER -