Hierarchical Linear Bayesian Method And Its Application In Haze Influencing Factors

Hierarchical linear model,which is also known as multi-level model,is a statistical model for the construction of nested structure data,and it is widely used in various fields such as socio-economic-ecology,the model is one of the most important statistical modeling methods of social science in the 21 st century.Bayesian statistics,which originated in the 18 th century,is a statistical discipline with significant status and value in today's big data era,and the research on its methods has permeated almost all fields.We apply bayesian statistics to the statistical inference of hierarchical linear model,which is an important research direction in current statistics field.At present,foreign scholars have conducted a series of studies on this field,but domestic scholars have relatively few researches on it.From the perspective of bayesian statistics,we explored the statistical inference of the hierarchical linear model and make an empirical analysis on the influence factors of haze in the hierarchical linear bayesian method.Firstly,this paper studied the bayesian statistical inference method of hierarchical linear model.This paper explored the prior selection method of hierarchical linear model.On the basis of describing the determination of prior distribution of the hierarchical linear model,this paper explored the priori choice and optimal problem.Secondly,the application of Gibbs sampling method is carried out for the posterior calculation of the hierarchical linear model.Thirdly,based on the statistical inference and model selection of the hierarchical linear bayesian model,a comparative study of the two methods of bayesian and empirical bayes is emphasized.Then,this paper constructed a three-level linear bayesian model and applied it to analyze the influence factors of haze.In the present and the future,smog problem is a major environmental problem,the monitoring data of the causes of haze is obviously nested,however,the traditional modeling method is difficult to effectively reflect this kind of data,which has a dependency structure.Therefore,we constructed a three-layered linear bayesian development model for haze influencing factors.After the inspection of empty model and based on the bayesian statistical inference,we carried out a comparative analysis of complete bayesian,empirical bayesian and classical maximum likelihood and limit maximum likelihood method,which highlighted the advantage of complete bayesian approach.On the basis of above,this paper provides some strategies and recommendations for more effectively control the haze.