Design of fuzzy systems using knowledge extracted via neural networks
Abstract
Automated identification of sleep stages is an extremely complex task due to several factors: recognizing and extracting relevant features of biomedical signals and variations of these features among subjects. Although conventional classification methods have proven to be effective, alternated methods utilizing fuzzy logic and neural networks offer an excellent option to classify sleep stages in a fast and reliable manner. The information utilized in this research was obtained from 8 rats sleep recording, which have been previously normalized by the expert and where only 3 stages are being classified: AWAKE, NON-REM and REM. In order to classify such stages, it was necessary to apply a knowledge extraction method to obtain the rules to base the fuzzy inference system. The results in the Huang and the other guy paper were improved for the Iris data classification from 86–91.33%, while the sleep stage classification was 95% accurate.
Subject Area
Computer science|Artificial intelligence
Recommended Citation
Martin Del Campo, Elvia, "Design of fuzzy systems using knowledge extracted via neural networks" (2004). ETD Collection for University of Texas, El Paso. AAIEP10796.
https://scholarworks.utep.edu/dissertations/AAIEP10796