Fuzzy Inference Based A Posterior Decision-Making for Multi-Objective Diet Optimization Problem

Authors : Cumali TÜRKMENOĞLU, Ayşe Şima UYAR, Berna KİRAZ

Pages : 41-47

Doi:10.31590/ejosat.1216648

View : 22 | Download : 15

Publication Date : 2022-12-31

Article Type : Research Paper

Abstract :We propose a Mamdani-Type Fuzzy Inference based posterior decision-making approach to multi-objective diet optimization problem. We optimize the multi-objective diet problem with evolutionary algorithms that result in tens/hundreds of non-dominated solutions which is too large to pick one of them by the decision-maker. Even though all the solutions are optimized for all the objectives simultaneously, not all objective functions may be equally important to a user and, also their importance may change for that user over time. Our main goal is to develop an applicable method for representing and incorporating a decision maker\`s insert ignore into journalissuearticles values(DM); instant preferences for objectives into decision-making stage. The FIS based decision making can guide users to decide on the most suitable menus. User\`s instant preferences for each objective form rule sets. Using Mamdani type FIS in the post-decision process of the multi-objective diet problem is a novel contribution. A desirability measure is calculated by using rule sets and membership functions considering the objective values, and based on the desirability measure the most preferred menuinsert ignore into journalissuearticles values(s); are provided to the user. Our method can direct the DM to the region of interest in the search space of the multi-objective diet problem. Thus, the daily menu suggestions become more applicable, practical, and desirable for the users.
Keywords : Fuzzy Sets, Fuzzy Inference Systems, Multi Objective optimization, Diet Optimization