Rule Generation for Hierarchical Fuzzy Systems (1997)
| Venue: | Proceedings of the Annual Conference of the North American Fuzzy Information Processing Society - NAFIPS ‘97, September 21 |
| Citations: | 3 - 0 self |
BibTeX
@INPROCEEDINGS{Holve97rulegeneration,
author = {Rainer Holve},
title = {Rule Generation for Hierarchical Fuzzy Systems},
booktitle = {Proceedings of the Annual Conference of the North American Fuzzy Information Processing Society - NAFIPS ‘97, September 21},
year = {1997},
pages = {24}
}
Share
 |
 |
 |
 |
||
OpenURL
Abstract
In this paper a new method of rule generation for hierarchical fuzzy systems (Hierarchical Fuzzy Associative Memory, HIFAM) is described. A HIFAM is structured as a binary tree and overcomes the exponential growth of the rulebases when the number of inputs increases. The training algorithm for HIFAM is suited for approximation and classification problems. Several benchmarks demonstrate that the proposed method compares well with existing learning techniques like artificial neural networks or decision trees. 1. Introduction The building of rule based function approximators and classifiers from training examples, that maintain a certain interpretability has gained growing interest in real world applications. The task is, to split the input space of a learning problem into several regions in order to separate the training examples (see Fig. 1): . Decision tree techniques like ID3 or C4.5 [1] use axis parallel splits to partition the input space. The positions of the splits are determine...
Keyphrases
hierarchical fuzzy system rule generation input space classification problem certain interpretability new method several benchmark real world application input increase hierarchical fuzzy associative memory exponential growth decision tree artificial neural network use axis parallel split learning problem binary tree training algorithm training example several region function approximators
