| Summary: | 碩士 === 高苑科技大學 === 資訊科技應用研究所 === 100 === In Taiwan, there are almost elections every year. No matter how the political dominions and administrative divisions change, or the election is intense, being a Taiwanese civil should select a candidate with morality and capability to promote the peoples’ benefit. Election competition is a zero-sum game with the result of either win or loss. This activity is also regarded as a complicated multi-criteria decision making problem. Traditionally, the election prediction is usually conducted by a public opinion poll. However, there are many problems of survey measurement, such as random sampling errors or people change their minds after they are surveyed; result in confusion and inaccurate information during a campaign period. This study therefore proposes a forecasting model based on the fuzzy preference relations under fuzzy environments to predict the chance of winning an election. The importance weight of evaluation criteria and candidates’ performance are determined by using linguistic terms parameterized in an interval [0, 1]. By multiplying the importance weights of evaluation criteria and the priority weight of the candidate, a degree of the candidate’s performance then will be obtained. This study uses the illustration to demonstrate the computation processes of this proposed approach.
|
|---|
