Application of The Fuzzy Grid Partition-Based Mamdani Method to Decision Support System for Determining Salary Increase Eligibility

Abstract

Fuzzy grid partition has been used to produce appropriate and optimal output. The output of several fuzzy inference methods such as the Tsukamoto method and the Mamdani method have been improved by applying grid partitions. This study aims to apply the fuzzy grid partition-based mamdani method to determine the feasibility of increasing employee salaries. The stages of the method are carried out starting with determining the number of partitions, forming fuzzy sets, carrying out the process of implicit rules, carrying out rule composition by selecting the maximum value of feasible and infeasible decisions, and finally carrying out the defuzzification process to obtain the calculation of the crips value. Attributes or features for selecting the eligibility for a salary increase are employee status, class status, years of service and benefits received. The results of the research get a decision for each employee whether it is appropriate to receive a salary increase or not. From one sample data tested on sample X7, the result of defuzzification (Z) is 5. Based on the table of feasible and inappropriate decisions, the value of Z = 5 is in a feasible decision