复杂系统多目标综合评价方法的比较研究

Comparative Research on Comprehensive Evaluation Approaches to Complex Multi-object System

主成分分析、模糊评判和数据包络分析是目前主要的综合评价方法,在实践中得到广泛应用。这些方法的原理各不相同,在介绍原理的基础上分析得出3种方法的适用性和局限性,针对各自的局限性指出相应的解决方法。结合粗糙集理论和k最邻近方法提出综合评价的混合算法,算法可以充分结合主客观因素,计算简单,在实际应用中具有可行性。

Nowadays, principal component analysis, fuzzy evaluation and DEA are main methods of comprehensive evaluation and have been applied extensively in many fields. These methods are different on the principle,therefore, have different applicabilities and limits. Paper brings forward the corresponding approachs of improvement based on analysis. Finally, bsaed on principle of Rough Set and k-nearest neighbors,paper puts forward the new combined algorithm of evaluation. Algorithm can well integrate subjective factors with objective factors and is simpler and more feasible in application.