Optimal Design of Rule-Based Systems by Solving Fuzzy Relational Equations
In this chapter, an approach to the design of rule-based systems within the framework of fuzzy relational calculus is proposed. The system of fuzzy relations serves as the generator of the rule-based solutions of fuzzy relational equations. Each solution represents a different trade-off between the classification accuracy and the number of fuzzy rules. The accuracy-complexity trade-off is achieved by optimization of the total number of decision classes for relations and rules.
- Peeva K, Kyosev Y (2004) Fuzzy relational calculus. Theory, applications and software. World Scientific, New YorkGoogle Scholar
- Rotshtein AP, Kuznetcov PD (1992) Design of faultless man-machine technologies. Technika, Кiev, Ukraine (in Russian)Google Scholar
- Rotshtein A, Rakytyanska H (2013) Expert rules refinement by solving fuzzy relational equations. In: Proceedings of IEEE conference on human system interaction, pp 257–264Google Scholar
- Yager RR, Filev DP (1994) Essentials of fuzzy modeling and control. Willey, New YorkGoogle Scholar
- Zhang D, Duan A, Fan Y, Wang Z (2008) A new approach to division of attribute space for SVR based classification rule extraction. Adv Neural Netw 5263:691–700Google Scholar