Research On Influencing Factors Of Guangxi Population Distribution Based On Random Forest Model And Geodetector

Revealing the pattern of population distribution and its influencing factors is an important guide for understanding the relationship between people and land,reconciling the contradictions between population,resources and environment in the region,and formulating sustainable development policies.This thesisn analyzes the sparse and clustered characteristics of the population distribution in Guangxi based on township scale,explores the factors affecting the population distribution in Guangxi and its regional differences,and spatializes the population data.A final comparison discusses the similarities and differences between the results of the random forest model and the geodetector on the exploration of factors influencing population distribution,the strengths and weaknesses of the approach.Using spatial statistical methods,stochastic forest models,geodetector models,etc.,the paper draws the following main conclusions.(1)The distribution of population density in Guangxi shows the characteristics of low in the northwest and high in the southeast,which roughly forms the town of Meixi in the northeast of Guilin City's Resources County.--Aidian Town in the southwest of Ningming County,Chongzuo City,is the population sparse dividing line,forming a population in Guigang,Yulin,Nanning,and Liuzhou High-density agglomerations have formed low-density population areas in Baise and Hechi.(2)The use of the stochastic forest model can be used to rank the importance of factors influencing the spatial differentiation of population density in Guangxi townships.The top five influences are distance from POI,nighttime light index,distance from road,man-made surface index,and road network density.The relationship between them and population degree is not simply linear,but non-linear or segmented.(3)There are significant differences in the spatial differentiation of each influence factor on the population density of townships in different economic subdivisions,with the Gui Dong area distinguished from the The main effects in other regions are road network density and average multi-year air temperature,the density of the river network in the western Guizhou region and the desertification index on population distribution.The intensity of impacts is higher than in other regions.The main factors of influence distinguishing the southern part of Gui from the other regions are the arable land index and vegetation cover.The main influencing factor in Guibei is the grass index,while the main influencing factor in Guizhong is the arable land index and gross industrial output.The population of the Guangxi region is mainly concentrated in the plains and basin areas,mainly in the Xunjiang Plain in Guigang,Guizhong in the Laibin area of Liuzhou Plain,the Right River Plain in Baise,the Nanning Basin,the Yulin Basin,and the Nanliu River Delta in the Beibu Gulf.(4)Ranking the importance of the results of the analysis of influences on population distribution using the geodetector model and the random forest model The results are similar,and the risk detection results are consistent with random forest zoning simulations,in the absence of finer demographics than at the township scale.Where the results of spatialization of random forest population data are validated,geoprobe models can be used for comparative analysis between models.By integrating the similarities and differences of the two influencing factors,we can better analyze the complexity of population distribution and build a population distribution system in Guangxi.Conceptual Model.