Title :
Spatial Combination Interpolation Model Based on Panel Data and Its Empirical Study
Author_Institution :
Sch. of Manage., Fuzhou Univ., Fuzhou, China
Abstract :
To discuss the spatial interpolation based on panel data with spatial autocorrelation, the first-order spatial autoregressive interpolation model and the Kriging algorithm interpolation model are established from the perspective of the cross-sectional data. Genetic algorithm back-propagation neural network interpolation model is established from the perspective of the time-series data. A spatial combination interpolation model is established by the results of these models. The weights of the combination model is calculated by a new method of spatial drift. An empirical study is carried out with interpolation some areaspsila GDP per capita in Fujian 2007, China. The result shows that the most effective one is the spatial combination interpolation model.
Keywords :
backpropagation; data analysis; genetic algorithms; interpolation; neural nets; time series; Kriging algorithm interpolation model; backpropagation neural network interpolation model; cross-sectional data; first-order spatial autoregressive interpolation model; genetic algorithm; panel data; spatial combination interpolation model; time-series data; Autocorrelation; Conference management; Economic indicators; Environmental management; Gallium nitride; Genetic algorithms; Interpolation; Neural networks; Reactive power; Technology management; panel data; spatial combination interpolation model; spatial drift;
Conference_Titel :
Environmental Science and Information Application Technology, 2009. ESIAT 2009. International Conference on
Conference_Location :
Wuhan
Print_ISBN :
978-0-7695-3682-8
DOI :
10.1109/ESIAT.2009.418
