Tournament Evaluation System Applying Win-Lose Result Presumption Considering Kansei Evaluation by Multiple People
We describe an Interactive Genetic Algorithm (IGA) with tournament evaluation for applying win-lose results obtained from an evaluation using multiple people. In our previous study, we developed an IGA with tournament evaluation as a basic model for evaluating candidate solutions using votes from multiple people [13]. However, tournament evaluation requires that IGA users evaluate the same candidate solution multiple times. Therefore, our previous method can reduce a user’s motivation for evaluating the solution. In addition, the number of users participating in a vote may decrease because of the decreasedmotivation to evaluate. To overcome this difficulty, we propose the application of a win-lose result presumption based on the tournament evaluation records of multiple people. When a system-based presumption is possible, the win-lose result presumption automatically determines the preferred and non-preferred candidates in each round. This method can reduce the number of times that users need to evaluate the same candidate solution. The effectiveness of the proposed method is verified using a numerical simulation that employs multiple numerical evaluation agents instead of human evaluators. The simulation results show an initial convergent improvement with the proposed method.