Multi-source data population spatialization method based on random forest-point-to-surface Kriging regression

Abstract

The invention relates to a multi-source data population spatialization method based on random forest point-to-surface Kriging regression, and belongs to the technical field of Internet and computers.The method specifically comprises the following steps: driving factor screening and data processing: preprocessing acquired geospatial data and social perception data, generating auxiliary data, and generating meshed covariable data from the auxiliary data; grid data upscaling aggregation: aggregating the gridded covariable data to a county-level administrative unit to generate a covariable on a population general survey unit scale; combining the population general survey data with county-level administrative unit data to obtain population density data of county-level administrative units; executing random forests on the population general survey data and the covariables, and then executing surface-to-point Kriging regression on residual components of the random forests; and combining thesurface-to-point Kriging regression result under the fine grid scale with the reserved random forest result to obtain a gridding population distribution diagram based on random forest point-to-surfaceKriging regression.

Description

technical field [0001] The invention belongs to the field of Internet and computer technology, and relates to a multi-source data population spatialization method based on random forest-point-to-surface kriging regression. Background technique [0002] The spatial distribution of population refers to the geographical distribution of population at a certain point in time. It is the spatial manifestation of the population process. It is the core issue of population geography research and an important basis for the study of human-land relationship. Data on population are generally demographic data describing the size, structure, and other information of the population within each defined statistical unit (such as administrative units, postal code areas, and census tracts). Census data is the main source of demographic data, which has certain limitations in geoscience applications. First, census data provide only one population count value per census unit; therefore, it does no...

AI Technical Summary

Problems solved by technology

Although many scholars have made a lot of explorations on different data sources, different scales, and different simulation methods, most of them are mesoscale studies with a spatial resolution of 1 km, which is difficult to meet the current requirements for refined urban management, and relatively The impact of spatial variables on population distribution is less explored

Images