Zinātniskās darbības atbalsta sistēma
Latviešu English

Publikācija: A Comparative Analysis of Prism and Mdtf Algorithms

Publikācijas veids Publikācijas konferenču materiālos, kas ir indeksēti Web of Science un/vai SCOPUS
Pamatdarbībai piesaistītais finansējums Nav zināms
Aizstāvēšana: ,
Publikācijas valoda English (en)
Nosaukums oriģinālvalodā A Comparative Analysis of Prism and Mdtf Algorithms
Pētniecības nozare 2. Inženierzinātnes un tehnoloģijas
Pētniecības apakšnozare 2.2. Elektrotehnika, elektronika, informācijas un komunikāciju tehnoloģijas
Autori Madara Gasparoviča-Asīte
Ludmila Aleksejeva
Atslēgas vārdi IF-THEN fuzzy classification rules; Machine learning; Membership functions; Modular rules; Decision table; Iris data; MPG data
Anotācija This research compares two algorithms that work with fuzzy data – the fuzzy PRISM algorithm and an algorithm for finding relevant attributes and membership functions – MDTF (merging – decision – table – first). Suitable data sets were searched for in the UCI Repository to examine the performance of both algorithms on real data. The most suitable data sets were Iris data set and Miles Per Gallon (MPG) data set. A series of experiments were carried out to compare the possibilities and working principles of each algorithm, depending on the size of the data set and the number of attributes used. As a result, recommendations for use of the algorithms are given, depending on the structure of the data set.
Atsauce Gasparoviča-Asīte, M., Aleksejeva, L. A Comparative Analysis of Prism and Mdtf Algorithms. No: Proceedings of 16th International Conference on Soft Computing (MENDEL 2010), Čehija, Brno, 23.-25. jūnijs, 2010. Brno: Brno University of Technology, 2010, 191.-197.lpp. ISBN 978-80-214-4120-0.
ID 7804