基于不变权重次约束的综合评价模型

大部分综合评价研究只解决了评价结果的排序问题,无法具体给出哪些因素才是影响具体评价对象特定评价结果的关键因素.次约束综合评价模型开拓了萃取影响评价对象关键因素的思路,但现有次约束研究的不足也很明显,将1个指标数据替换为最差值0,即去掉了这个指标的贡献后的评价得分反而可能会高于该指标得分不为0时的得分,这就是"完整指标体系"的评价得分反而小于"一个或几个指标对得分的贡献被去掉后"的评价得分的不合理现象.文章的创新与特色一是通过保持有n个指标的完整指标体系与n-1个指标的指标体系前后两次评价中指标的权重不变,来改进现有的次约束模型变权重的做法,改正"完整指标体系"的评价得分反而小于"一个或几个指标对得分的贡献被去掉后"的评价得分这种不合理现象.二是通过改进评价标准,在评价不同对象并萃取其优势和劣势因素时,将1个指标得分变为0前、后评价得分的排序是否变化的标准,改进为比较评价得分变化量大小的标准,避免了因前、后两次评价得分排序的序关系不变而导致的无法萃取优势或劣势因素的弊端.三是通过比较将1个指标得分变为0前、后两次评价得分变化量的大小,萃取出影响副省级城市经济社会发展的关键因素、优势和劣势因素.

A comprehensive evaluation model of worst constraint based on constant weight

Comprehensive evaluation models has broad applications in the process of economic and social de-velopment.Most existing studies can only work out the rank of the results ( which result is the best, which is the second best, and so on), without showing which factors are the key factors resulting in the rank.The Worst Constraint Model can extract the key factors by comparing the evaluation score made by the complete in-dex system with the score made by the incomplete index system in which the value of some indices are turned into zero.The score derived from the complete index system can be lower than the score derived from the in-complete index system in Worst Constraint Model, i.e., the score of the complete index system can be lower than the score of the index system.This paper improves the Worst Constraint Model by keeping the weight of the index constant instead of re-weighting the indices whose values do not turn into zero.So, this improved model will correct the unreasonable error in the Worst Constraint Model.The special contribution of this paper lies in three aspects.First, the weights of index are kept constant, which means the complete index system has the same weight, instead of re-weighting the indices whose values do not turn into zero.This can ensure"the score of the complete index system must be higher than the score of the index system".Second, the cri-teria of extracting dominant factors and restraining factors are improved from"comparing the score rank of the complete index system and that of the incomplete index system" to "comparing the change in the quantity of score of the incomplete index system and that of the complete index system".This improvement solves the problem that existing researches can not extract dominant factors and restraining factors when the rank is un-changed.Third, the key factors, dominant factors, and restraining factors that influence the economic and so-cial development of 15 deputy provincial cities are extracted using the improved Worst Constraint Model and improved criteria of extracting dominant factors and restraining factors.