Work

An Approach to Generating Fuzzy Rules for a Fuzzy Controller Based on the Decision Tree Interpretation

Year: 2025

Type: article

Abstract: The article considers the solution of control problems based on fuzzy logic. This approach allows to build decision support systems in various domains. The novelty of the approach is the algorithm for the generation of the fuzzy rules for a fuzzy controller based on the machine learning results interpretation to improve the quality of control actions in organizational and technical systems. Machine learning methods can find unknown patterns that require deep expert knowledge in some domain with a manual rules construction. We consider an example of the generation of a set of fuzzy rules based on the analysis of a decision tree model. It is possible to generate a set of fuzzy rules for constructing fuzzy inference system (FIS) based on the proposed algorithm. Membership functions and labels of linguistic terms are generated automatically for all input and output variables. The quality of the machine learning model and FIS were evaluated using the R2 metric. Experimental testing showed what the quality of FIS that is based on the generated fuzzy rules is worse by an average of 2 % compared to the original model based on the decision tree. The quality of FIS can be improved by tuning the membership functions, but this issue is beyond the scope of this article.

Source: Axioms

DOI: 10.3390/axioms14030196

Authors: Anton A Romanov, Aleksey A Filippov, Nadezhda G Yarushkina

Institution: Ulyanovsk State Technical University

Language: English

Topic: Fuzzy Logic and Control Systems

Subfield: Artificial Intelligence

Field: Computer Science

Domain: Physical Sciences

Open Access status: gold

Estimated APC paid: $2,896

Funder: Ministry of Science and Higher Education of the Russian Federation