Font Size: a A A

The Control And Distribution Of The Collages Assets Based On Bayesian Estimation

Posted on:2017-02-12Degree:MasterType:Thesis
Country:ChinaCandidate:S Q YuFull Text:PDF
GTID:2297330503985509Subject:Probability theory and mathematical statistics
Abstract/Summary:PDF Full Text Request
How to make a reasonable criterion to promote the construction and development of colleges and universities is an urgent practical issue for asset demand varies in different colleges and universities.This paper research the changes in assets by applying Bayesian estimation method. Then based on Bayesian statistics method and prediction posteriori distribution, a reasonable criterion which is suitable for the development of colleges and universities, is given to measure the development of them and control their assets reasonably and effectively.The main research of this paper is as follows:(1) In order to get the law of parameters in asset index variable distribution, asset index variable distribution is studied dynamically by Bayesian estimation. Then according to the obtained distribution, assets are estimated and controlled. A hierarchical Bayes model is established to study the asset index variable distribution dynamically for variables which after the logarithmic function transformation are obey the normal distribution and the results show that asset index variable distribution has a small fluctuation.(2) The prediction posteriori distribution of asset index variable is obtained by employing the asset distribution and posterior distribution. Then based on the prediction posteriori distribution, changes about the assets are predicted. The predicted results not only illustrate the effect of changes in historical and new asset samples but also display the improvement of assets estimation under the unknown assets criterion.(3) Further, measurement of an asset belonging to different levels by establishing BP neural network model and defining the mixed membership. Though the prediction accuracy reaches 94.7%, this model depends on samples under the given asset criterion. Therefore, compare to Bayesian estimation, there are still some deficiencies.(4) Solve the issues of assets and assets control by applying the Bayesian estimation and BP neural network model.
Keywords/Search Tags:Asset Distribution, Bayesian Statistics, Mixed Membership, Posteriori Prediction Distribution, Prediction and Control
PDF Full Text Request
Related items