基于正负理想点的评估指标权重选取方法

Weight Determination Approach for Assessment Index Based on Positive and Negative Ideal Points

针对现有方法在评估指标权重选取上无法兼顾主客观2种因素的缺陷,提出基于正负理想点的指标权重选取方法。该方法仅需一个样本集以及从中选定的最优与最差评估对象,便可基于这些信息构造代价函数,并通过模拟退火优化过程求出指标权重。由于代价函数有效整合了主观和客观信息,故两者自然体现于权重结果中,并进行实验分析。结果表明:该方法能在统一的计算过程中解决主客观因素的融合问题,为评估指标权重的选取提供新的思路和途径,明显优于现有方法。

Aiming at the defect of the existing methods in failing to take into consideration both of the objective and subjective factors for index weight determination, this paper proposed a weight determination method based on positive and negative ideal points. This method only needed a sample set as well as the best and worst assessment objects selected from the set, then a cost function could be built on the given information and thus the index weight could be calculated through the simulated annealing optimization process. The objective and subjective information would be naturally reflected in the calculated weight, due to the effective integration of them both into the cost function, then an experiment analysis was conducted. The experimental results show that this method is superior to its existing counterparts because it solves the problem of the integration of the objective and subjective factors in a unified calculation process, offering new ideas and approaches for index weight determination.