SyFSeL: generating synthetic fuzzy sets made simple

Empirical tests can help determine if methods developed for fuzzy sets work correctly. However, finding a large enough data set with suitable properties to conduct thorough tests can be challenging. This paper presents a new library named SyFSeL (Synthetic Fuzzy Set Library) which automatically gene...

Full description

Bibliographic Details
Main Author: McCulloch, Josie
Format: Conference or Workshop Item
Language:English
Published: 2018
Online Access:https://eprints.nottingham.ac.uk/53092/
Description
Summary:Empirical tests can help determine if methods developed for fuzzy sets work correctly. However, finding a large enough data set with suitable properties to conduct thorough tests can be challenging. This paper presents a new library named SyFSeL (Synthetic Fuzzy Set Library) which automatically generates synthetic fuzzy sets with specified characteristics and fuzzy set type. SyFSeL generates as many sets as desired, with adjustable parameters to enable users to emulate real data. Generated fuzzy sets are exported so users can import them into their own fuzzy systems software. SyFSeL can also create graphical plots of the generated sets, examples of which are shown in this paper. The library is cross-platform and open-source under the GNU General Public License, and users are free to develop upon and adapt the code. However, SyFSeL has been designed so that no understanding of the code is required to use it.