Research On Bayesian Estimation Of Risk Of Resource Allocation Of Universities And Control Model

According to the risk problem of optimization allocation of university resources,based on the collection of the performance of input and output data of 72 universities in 5 years,according to the deviation between performance and the average performance,the deviation between the performance and the maximum performance,and the possibility of two deviations as risk factors,BP neural network is used to build the nonlinear relationship between risk level and influencing factors.Then explore Bayesian estimation of risk and evaluate the risk of resource allocation process.In order to avoid the irrationality or loss of resource allocation as far as possible,the risk control is considered in a reasonable range.For the specified variable range of the risk factors,seek the variable range of variables of influence factors of risk.The control model of risk is established by genetic algorithm to explore the change territory of the corresponding risk factors when the risk random variables change.Unlike the traditional genetic algorithm model,the genetic algorithm model of this paper can get a control field with prediction interval,which is indefinite,more in line with the characteristics of risk in real life.These methods can not only manage and monitor the resource allocation process,but also improve the efficiency of the use of resources better.The results of this paper provide an effective evaluation and control method for the risk problem of resource allocation in colleges and universities.The main work of this paper is as follows:(1)The risk is defined as a random variable under the influence of multiple factors random variables and their occurrence.(2)The K-Means clustering algorithm is used to obtain the comprehensive level of risk,and then the BP neural network is used to construct the nonlinear relationship f between the risk level and the influencing factors.(3)Bayesian method is used to explore the prior distribution and posterior distribution of risk as the risk is taken as a random variable.Then Bayesian estimation of the risk of 72 colleges and universities in 2007-2011 is obtained.The estimation error is also compared.(4)The risk control model is established.When the risk changes as a random variable with a certain probability in the expected range,genetic algorithm is applied to solve the value range of risk variables.Finally the control domain under known is tested and analyzed.