Based on the evaluation and sorting of enterprise's technology innovation ability, this paper compares the reliability of measuring weight of evaluating index by Delphi method, Analytic Hierarchy Process(AHP), Entropy method and Fuzzy Cluster Analysis, to research applicable area and advantage and disadvantage of each method. The paper draws conclusion through theoretical analysis especially practical compare that if complete sample datum can be available, the reliability of Entropy method is higher. If sample datum is scare, the reliability of AHP is higher than Delphi method. If there are a lot of fuzzy indexes among sample, we should adopt Fuzzy Cluster Analysis, then conclude the weights of indexes by AHP.%以企业技术创新能力评价排序问题为实证范例，对德尔菲法、层次分析法、熵值法和模糊聚类分析法测定评价指标权重的可信度做了比较研究，目的是探索各自的适用范围和优缺点．文献通过理论分析特别是实证比较，认为样本数据完整，熵值法则较高；如果缺乏样本数据，则应采用AHP法，可信度高于德尔菲法；若样本中含有大量的模糊数据，且同一层次指标个数较多，应先采用模糊聚类分析法分类，再用AHP法分别做各个子类的权重．
展开▼
