| The theory of spatial statistics first developed in 70 years, so we can find that spatial statistics is a new subject. Spatial statistics is different from classical statistics. Many theories of spatial statistic root in classical statistics, but the analytical method of the model of spatial statistics and the building of the model are dissimilarity to classical statistics. The major character of spatial statistics is that it has a"space"concept. Because we can not leave from the abstract and material spatial concept.Spatial statistics deals with spatial data. Spatial data has location geographical attribute, it differs from cross-sectional and time series data. We always use the spatial method to study the point in the space. We study the point in the space whether has the character of aggregation, and how to measure the intensity of aggregation. The most particularity of spatial statistics is that it has a new concept of"proximity", it brings the proximity to other scopes of statistics. So it gives birth to new statistical models, spatial models. For example, y =λW y + Xβ+ε.The model is different from classical statistics model. The research work of spatial statistics is to estimate parameters of spatial model. the method of estimation of spatial model is not perfect by now, especially for the spatial model with missing data.The reason of developing spatial statistics is that it has many applications in our life, and the model has considered the locations. Our activities must occur in the geographical locations. So the spatial model can better study some problems. For example, the analysis of spatial statistics of china regional economic.In the article, there is some following research work: First, estimate the parameter of the special spatial model and discuss the asymptotical property of parameters; Second, discuss the spatial model with missing data; In the end, use spatial statistics to study regional economic and put forward some suggestions. |
PDF Full Text Request