作者: Pavel Kromer , Jan Platos , Vaclav Snasel
DOI: 10.1109/CYBCONF.2013.6617453
关键词:
摘要: Methods based on fuzzy sets and logic have proved to be efficient data classifiers value estimators. This study presents an application of evolutionary evolved rules the concept extended Boolean queries a multi-class mining problem. Fuzzy are used as symbolic machine-learned from label samples predict output variable. The variable can both (category) continuous value. prediction quality attributes in industrial set compares obtained by achieved support vector machines.