@inbook{41d3b5185dec4fff97d068640cea7017,
title = "Application of Rough Set-Based Characterisation of Attributes in Feature Selection and Reduction",
abstract = "Quality of predictions depends heavily on features that are chosen for a classification system to rely on. It is one of the reasons why approaches, focused on feature selection and reduction, play a significant role in data mining. Among all available attributes, these should be detected that are of the highest relevance and importance for a given task. This objective can be achieved by an application of one of feature ranking algorithms. Some of data exploration methods have their own inherent mechanisms dedicated to feature reduction, and decision reducts, defined within rough set theory, offer such option. The chapter presents research on application of reduct-based characterisation of features, employed to support classification by selected inducers working outside rough set domain. The problem to be solved comes from the field of stylometry. It is the study of writing styles with the main task of authorship attribution, while using characteristic features not of qualitative, but quantitative type.",
keywords = "Authorship attribution, Classification, Decision reduct, Feature selection, Ranking, Rough set theory, Stylometry",
author = "Urszula Sta{\'n}czyk",
note = "Publisher Copyright: {\textcopyright} 2022, The Author(s), under exclusive license to Springer Nature Switzerland AG.",
year = "2022",
doi = "10.1007/978-3-030-93052-3\_3",
language = "English",
series = "Learning and Analytics in Intelligent Systems",
publisher = "Springer Nature",
pages = "35--55",
booktitle = "Learning and Analytics in Intelligent Systems",
address = "Switzerland",
}