AN INFORMATION THEORETIC APPROACH TO THE ESTIMATION OF MIGRATION FLOWS*
比较了人口迁移建模的几种方法，提出扩展的最小信息模型最适合利用美国内部迁移数据构建动态空间交互模型。
This study is concerned with the application of different types of models to the study of migration flows. Two traditional approaches to migration modeling, the demographic fixed-rate and the gravity/entropy families of models, are first described, and their adequacy is assessed. The author next discusses the derivation of minimum information models. Consideration is given to destination-weighted models and to net-constrained models. A unified approach that incorporates both a priori micro and estimated macro parameters is then considered. The author concludes that the extended minimum information approach is the most suitable for building dynamic spatial interaction models, given the current state of U.S. internal migration data.