Analysis Of Sci-technology Innovation Ability Of Universities Based On The Combination Of Cloud Model And Threshold Model

IIUniversities are one of the important innovation bodies,and their scientific research innovation level directly affects the national scientific and technological development capability.Affected by factors such as region,history,and subject development,there are differences in sci-technology innovation ability among universities.Scientific evaluation of the sci-technology innovation ability of universities is conducive to helping them improve their sci-technology innovation ability.The existing research on the sci-technology innovation ability of universities has achieved certain results in the methods of evaluation and analysis of influencing factors.However,there are still problems such as the lack of consideration of the geographical location and the differences in the characteristics of disciplines and majors of various universities,as well as the inconsistent evaluation indicators and the subjectivity of weight determination.As a result,the evaluation results are not objective and comprehensive;and there is less discussion on the nonlinear relationship in the analysis of the influencing factors of innovation ability.Therefore,it is of great significance to explore more scientific and reasonable comprehensive evaluation methods and influencing factor analysis methods.This paper specifically carried out the following research work: First,the cloud model and the threshold model were constructed to complete the evaluation and influencing factor analysis of the sci-technology innovation ability of universities;Second,the data of the sci-technology innovation activities of universities directly under the Ministry of Education from 2009 to 2016 were used for empirical analysis.Third,the evaluation index system was established from six aspects: innovation environment,human input,financial input,international cooperation and innovation ability,output ability and scientific and technological achievements transformation ability.Fourth,this thesis made an empirical study on the comprehensive evaluation and influencing factors of sci-technology innovation ability of universities in different types and regions.The research characteristics as follows: First,introduced new comprehensive evaluation methods and influencing factor analysis methods in the field of scitechnology innovation ability in universities.Second,the classification method of cloud model was optimized by cluster analysis method;the importance ratio of each index was obtained as weight by using random forest classification model,and the index weight of cloud model was optimized.This can not only improve the operability of the model,but also ensure the scientific evaluation.Research conclusions: First,the classification of universities by region and discipline type solved the incomparability problem caused by different regions and disciplines;the evaluation model established by the cloud model solved the problems of ambiguity in evaluation and the subjectivity of index weight determination;the establishment of the threshold model made up for the insufficient discussion of nonlinear relationship in the analysis of the influencing factors of sci-technological innovation ability of universities in the existing research.Second,the evaluation method of clustering method and cloud model combination has feasibility and obvious advantages in the evaluation of sci-technology innovation ability of universities.Third,the fitting effect of threshold model on panel data with threshold value is better than that of ordinary panel data regression model.