Gen*: a generic toolkit to generate spatially explicit synthetic populations

被引：19

作者：

Chapuis, Kevin ^{[1
]}

Taillandier, Patrick ^{[2
]}

Renaud, Misslin ^{[3
]}

Drogoul, Alexis ^{[4
]}

机构：

[1] IRD UPMC, UMI UMMISCO 209, Bondy, France

[2] Univ Toulouse, INRA, MIAT, Castanet Tolosan, France

[3] Univ Rouen, Identite & Differenciat Espace Environm & Soc, Mont St Aignan, France

[4] VAST, USTH, ICTLab, Hanoi, Vietnam

来源：

INTERNATIONAL JOURNAL OF GEOGRAPHICAL INFORMATION SCIENCE | 2018年 / 32卷 / 06期

关键词：

Synthetic population; spatialization; social simulation; multi-agent model; ETM PLUS IMAGERY; LAND-COVER;

D O I：

10.1080/13658816.2018.1440563

中图分类号：

TP [自动化技术、计算机技术];

学科分类号：

0812 ;

摘要：

Agent-based models tend to integrate more and more data that can deeply impact their outcomes. Among these data, the ones that deal with agent attributes and localization are particularly important, but are very difficult to collect. In order to tackle this issue, we propose a complete generic toolkit called Gen* dedicated to generating spatially explicit synthetic populations from global (census and GIS) data. This article focuses on the localization methods provided by Gen* that are based on regression, geometrical constraints and spatial distributions. The toolkit is applied for a case study concerning the generation of the population of Rouen (France) and shows the capabilities of Gen* regarding population spatialization.

引用

页码：1194 / 1210

页数：17

共 30 条

[1] Methods for Estimating Population Density in Data-Limited Areas: Evaluating Regression and Tree-Based Models in Peru [J].