Constrainedness In Stable Matching

2018 IEEE 30TH INTERNATIONAL CONFERENCE ON TOOLS WITH ARTIFICIAL INTELLIGENCE (ICTAI)(2018)

引用 0|浏览43
暂无评分
摘要
In constraint satisfaction problems, constrainedness provides a way to predict the number of solutions: for instances of a same size, the number of constraints is inversely correlated with the number of solutions. However, there is no obvious equivalent metric for stable matching problems. We introduce the contrarian score, a simple metric that is to matching problems what con strainedness is to constraint satisfaction problems. In addition to comparing the contrarian score against other potential tightness metrics, we test it for different instance sizes as well as extremely distinct versions of the stable matching problem. In all cases, we find that the correlation between contrarian score and number of solutions is very strong.
更多
查看译文
关键词
stable matching,constrainedness,contrarian score
AI 理解论文
溯源树
样例
生成溯源树,研究论文发展脉络
Chat Paper
正在生成论文摘要