Automatic Evaluation Based on Computer Science
Bo Li · Advanced materials research · 2014
In this paper, we develop an automatic system which could help us to select or evaluate the most suitable target from a variety of candidates. Above all, we carry out an analytic hierarchy process including 3 levels with different factors. After combing with fuzzy comprehensive evaluation, we rank the candidates with final scores. More specifically, taking the selection of the best coaches of football in American universities as an example, a hierarchical structure it built, with coaches’ comprehensive capacity being the target level, their technical skills, competition records, and personal influence being criteria level, coaching experience, teams’ improvement factors, win-loss percentage, participation in highly-demanding competitions and awards of coaches being the corresponding index level. Then, we use cluster analysis to determine threshold of each index, and combine trapezoidal distribution function with fuzzy comprehensive evaluation to obtain the membership degree of the following five sectors: fully equipped, equipped, basically equipped, poorly equipped, not at all equipped. After scoring each coach, we obtain the final rank