Comprehensive evaluation of universities scientific research capacity based on method set

Research capacity is a core competence of university, but also reflects an important indicator of the strength of university. Based on specific content of university research capacity, I analysis the composition of university research capacity, and propose the research capacity evaluation indicators of universities. In order to scientifically evaluate universities research capacity, this paper proposes a comprehensive evaluation based on PCA, BP neural network, fuzzy comprehensive evaluation, TOPSIS and DEA. Firstly, the model uses KENDALL-W correlation to check up the coherence of different method. Then using mean, Borda and Copeland model gets the comprehensive results. Subsequently, in order to evaluate the availability of results, Spearman correlation is used. And based on the Spearman correlation, the best model is selected. At last, through evaluating higher education research capacity of 20 universities proved the proposed model.