Concepts of matching are introduced by building three matched samples for the binge drinking example. Each matched sample is a solution to an optimization problem, and each is implemented by finding a minimum cost flow in a network. The third matched sample is judged the best of these three: It pairs closely for important covariates and balances other covariates.

错误:搜索内容不能为空,请输入英文关键词
错误:关键词超出字数限制,请精简
高级检索

Algorithms for Matching

  • Paul R. Rosenbaum

摘要

Concepts of matching are introduced by building three matched samples for the binge drinking example. Each matched sample is a solution to an optimization problem, and each is implemented by finding a minimum cost flow in a network. The third matched sample is judged the best of these three: It pairs closely for important covariates and balances other covariates.