CONSTRUCTION OF A FUZZY CLASSIFIER STRUCTURE BY COMBINING THE ALGORITHM BASED ON EXTREME VALUES OF FEATURES AND THE SHUFFLED FROG LEAPING ALGORITHM FOR IMBALANCED DATA WITH TWO CLASSES
M. B. Bardamova, I. A. Hodashinsky
Tomsk State University of Control Systems and Radioelectronics, Tomsk, Russia
Keywords: fuzzy classifier, shuffled frog leaping algorithm, fuzzy classifier structure, imbalanced data
Abstract
The paper proposes a way to apply the shuffled frog leaping algorithm as a tool to expand a rule base of the fuzzy classifier. Its application is relevant in the case where the existing rules are not sufficient for the qualitative recognition of all classes, for example, in the presence of data imbalance. Additional rules generated by metaheuristics are able not only to improve the classification quality, but also to provide a more complete description of the domain under study. To create a compact initial structure of the classifier, the algorithm based on extreme values of features in classes is used. The combination under consideration is tested on 36 imbalanced datasets from the Knowledge Extraction based on Evolutionary Learning repository and showed an increase in the mean geometric accuracy on 34 datasets, as well as satisfactory results compared to analogs. The advantages of the proposed algorithm of structure formation are the absence of the necessity of data augmenting with synthetic samples, low scatter of results in individual runs, and the ability to improve the classification quality by adding a few rules.
|
