@inbook{424c8272ae90419db34bf00b016cf436,
title = "Improving the quality of clustering-based diagnostic rules by lowering dimension of the cluster prototypes",
abstract = "The study concerns an evaluation of rule premise conditions generated from cluster prototypes. Different clustering methods are used as the first step of training the rule-based classifiers. If rules are generated directly from cluster prototypes, their premise complexity (the number of conditions) is equal to the dimension of considered training data. This paper describes an idea of detecting an appropriate part of the clustering-based rule premise and refinement of the rule set. It is especially valuable in medical domain applications of rule-based classifiers in the view of their generalization quality and interpretability.",
keywords = "Clustering, Fuzzy rule classifiers, Rule extraction, Rule set tuning",
author = "Sebastian Porebski and Ewa Straszecka",
note = "Publisher Copyright: {\textcopyright} Springer Nature Switzerland AG 2020.",
year = "2020",
doi = "10.1007/978-3-030-19738-4\_6",
language = "English",
series = "Advances in Intelligent Systems and Computing",
publisher = "Springer Verlag",
pages = "47--56",
booktitle = "Advances in Intelligent Systems and Computing",
address = "Germany",
}