Saved in:
| Main Author: | |
|---|---|
| Format: | Artículo científico |
| Language: | es |
| Published: |
Asociación Española para la Inteligencia Artificial
2006
|
| Subjects: | |
| Online Access: | https://www.redalyc.org/articulo.oa?id=92502912 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1866587796721893376 |
|---|---|
| author | V. Estruch |
| author_facet | V. Estruch |
| contents | Similarity Functions for Structured Data.An Application to Decision Trees V. Estruch C. Ferri J. Hernández-Orallo M. J. Ramírez-Quintana Ingeniería ILP Distance based Methods Decision Trees Structured Data Learning from structured data is becoming increasingly important. Besides the well-known approaches whichdeal directly with complex data representations (inductive logic programming and multi-relational data mining),new techniques have been recently proposed by upgrading propositional learning algorithms. Focusingon distance-based methods, these techniques are extended by incorporating similarity functions defined overstructured domains, for instance a k-NN algorithm solving a graph classification problem. Since a measurebetween objects is the essential component for this kind of methods, this paper starts with a description ofsome of the recent similarity functions defined over common structured data (lists, sets, terms, etc.). However,many of the most common classification techniques, such as decision tree learning, are not distance-basedmethods or cannot be directly adapted to be so (as kernel methods and neural networks have been adapted).In this work, we extend decision trees to use any kind of similarity function. The method is inspired bycentre splitting, which constructs decision trees by defining splits based on the distance to two or morecentroids. We include an experimental analysis with both propositional data and complex data. Apart fromthe advantages of the new proposed method, it can be used as an example of how other partition-basedmethods can be adapted to deal with distances and, hence, with structured data. 2006 artículo científico 1137-3601 https://www.redalyc.org/articulo.oa?id=92502912 es http://www.redalyc.org/revista.oa?id=925 Inteligencia Artificial. Revista Iberoamericana de Inteligencia Artificial application/pdf Asociación Española para la Inteligencia Artificial Inteligencia Artificial. Revista Iberoamericana de Inteligencia Artificial (España) Num.29 Vol.10 |
| format | Artículo científico |
| id | redalyc_92502912 |
| language | es |
| publishDate | 2006 |
| publisher | Asociación Española para la Inteligencia Artificial |
| spellingShingle | Similarity Functions for Structured Data.An Application to Decision Trees V. Estruch Ingeniería ILP Distance based Methods Decision Trees Structured Data Similarity Functions for Structured Data.An Application to Decision Trees V. Estruch C. Ferri J. Hernández-Orallo M. J. Ramírez-Quintana Ingeniería ILP Distance based Methods Decision Trees Structured Data Learning from structured data is becoming increasingly important. Besides the well-known approaches whichdeal directly with complex data representations (inductive logic programming and multi-relational data mining),new techniques have been recently proposed by upgrading propositional learning algorithms. Focusingon distance-based methods, these techniques are extended by incorporating similarity functions defined overstructured domains, for instance a k-NN algorithm solving a graph classification problem. Since a measurebetween objects is the essential component for this kind of methods, this paper starts with a description ofsome of the recent similarity functions defined over common structured data (lists, sets, terms, etc.). However,many of the most common classification techniques, such as decision tree learning, are not distance-basedmethods or cannot be directly adapted to be so (as kernel methods and neural networks have been adapted).In this work, we extend decision trees to use any kind of similarity function. The method is inspired bycentre splitting, which constructs decision trees by defining splits based on the distance to two or morecentroids. We include an experimental analysis with both propositional data and complex data. Apart fromthe advantages of the new proposed method, it can be used as an example of how other partition-basedmethods can be adapted to deal with distances and, hence, with structured data. 2006 artículo científico 1137-3601 https://www.redalyc.org/articulo.oa?id=92502912 es http://www.redalyc.org/revista.oa?id=925 Inteligencia Artificial. Revista Iberoamericana de Inteligencia Artificial application/pdf Asociación Española para la Inteligencia Artificial Inteligencia Artificial. Revista Iberoamericana de Inteligencia Artificial (España) Num.29 Vol.10 |
| title | Similarity Functions for Structured Data.An Application to Decision Trees |
| topic | Ingeniería ILP Distance based Methods Decision Trees Structured Data |
| url | https://www.redalyc.org/articulo.oa?id=92502912 |