Abstract:

Spatial analysis is a method of quantitative analysis of spatial phenomena, which becomes the core competitiveness of supporting the development of GIS. Especially with the rapid development of smart phones, mobile payment systems and shared bicycles, spatial analysis has been brought to a new stage of artificial intelligence in just a few years. In order to draw a blueprint for the coming artificial intelligence spatial analysis (AISA), this paper first briefly combs the key nodes of the development of spatial analysis: the computerized work mode from 0 to 1, global visual calculation, hidden LBS and intelligent applications, cloud GIS and
artificial intelligence spatial-temporal decision-making. Secondly, the five schools of machine learning, the evolution characteristics of dominant algorithms and spatial analysis are summarized. Thirdly, the definition, modeling principles and technical framework of AISA are proposed. Finally, several hot area of AISA in the future are proposed, such as intelligent spatial computing, hyperparametric spatial optimization, intelligent spatial planning robot, full-sample spatial-temporal prediction and spatial neural network analysis. By combing, analyzing, summarizing and predicting the history and trend of artificial inerlligence spatial analysis, this paper aims to provide systematic theoretical and applied data for artificial intelligence spatial analysis.