摘要
空间分析是对空间现象进行定量分析的一种方法,成为支撑地理信息系统(geographic information system,GIS)发展的核心竞争力。LBS(location based service)的非专业化应用,给空间分析带来了空前的挑战。尤其是智能手机、移动支付系统、共享单车等的快速发展,在短短几年时间内将空间分析带入到一个全新的人工智能阶段。为了给正在到来的人工智能空间分析(artificial intelligence spatial analysis,AISA)绘制一个蓝图,首先简要梳理了空间分析发展的关键节点:从0到1的计算机化工作模式、全球视野的可视化计算、隐LBS与智慧应用、云GIS与人工智能时空决策。其次梳理和总结了机器学习的五大流派、主导算法与空间分析对应的演进特点。再次,提出了人工智能空间分析的定义和建模原理、技术框架。最后,预测了未来人工智能空间分析的热点研究方向:智能空间计算、超参数空间优化、智能空间规划机器人、全样本时空预测和空间神经网络分析等。通过梳理、分析、总结及预测人工智能空间分析的发展历史及发展趋势,旨在为人工智能空间分析提供系统性的理论及应用研究参考。
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, hyper-parametric 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.
作者
郑敏睿
郑新奇
王娇
ZHENG Minrui;ZHENG Xinqi;WANG Jiao(Center for Applied GI Science,University of North Carolina at Charlotte,Charlotte,NC 28223,USA;School of Information engineering,China University of Geosciences（Beijing）,Beijing 100083,China)
出处
《自然杂志》
2018年第5期363-372,共10页
Chinese Journal of Nature
基金
国家自然科学基金项目(71673256
41801361)
国家国际科技合作专项项目(2015DFA01370)资助
关键词
AISA技术原理
超参数空间优化
空间神经网络分析
AISA technology principle
super parameter spatial optimization
spatial neural network analysis
