基于降维思想的客观组合评价模型

Objective Combined Evaluation Model Based on Dimension Reduction

为了解决多方法评价结论的非一致性问题,产生了组合评价方法。最大限度地利用多评价结论的信息是组合评价的关键。为此提出了一个基于降维思想的客观组合评价模型,它能够最大程度地保留多评价结论所包含的评价信息,充分地整合多评价结论中的共性信息。鉴于该模型求解属复杂非线性优化问题,常规方法难以直接处理,建立了微粒群改进算法,进行全局寻优。最后,通过实例说明了模型的可操作性和实用性。

In order to solve the inconsistency among the evaluation conclusions drawn by different evaluation methods, combined evaluation methods have emerged. How to maximize the utility of conclusion generated from multi-evaluation is the key to combined evaluation. Aimed at this key, an objective combined evaluation model based on dimension reduction is proposed. However, conventional method can hardly handle the complicate problem of non-linear optimization. In the light of this drawback, an improved particle swarm optimization algorithm to make a global optimization is established. This new combined evaluation model is superior to others in maintaining the conclusive information to the highest degree and integrating the common information to the fullest extent. Moreover, this model is characterized by its clarity, conciseness and easy implementation. Eventually, a case study illustrates the practicality and validity of this model.