Study on unascertained comprehensive evaluation model basing on modified entropy

Unascertained theory and method are mathematical tool to deal with uncertain information,which has rigorous logic. Traditional unascertained evaluation model determines index weight by information entropy method usually, in which the weight value is more objective, but information entropy method can´t express the knowledge and experience of experts and the opinion of decision makers, and sometimes obvious deviation may exist between the weight by entropy and the actual important degree, and even may lead to incorrect weight. This paper introduces the analytic hierarchy process to modify information entropy,and establishes the unascertained measure model basing on modified entropy weight. The model combines the weight determined by experts and by entropy weight to solve the index weight, which gives a comprehensive consideration of subjective and objective factors, and the method can improve the accuracy of weight value and the reliability of evaluation results.