代价敏感粗糙集研究综述

A Survey on Cost-sensitive Rough Set Research

本文对代价敏感粗糙集(cost-sensitive rough set,CS-RS)理论的内容及发展状况进行了回顾,对其研究方向和发展趋势进行了分析和展望.首先介绍了代价敏感学习产生的缘由、数据挖掘过程中常见代价的分类及代价敏感粗糙集的基本概念;其次从数据模型、计算模型、问题及算法四个层次上对当前的研究状况进行了总结,建立了代价敏感粗糙集理论体系示意图;最后指出测试代价敏感粗糙集的发展趋势和需要重点关注的问题.

This paper reviews the theory foundations and the state-of-the-art of cost-sensitive rough set. Firstly, the motivation of cost-sensitive learning is introduced through examples. Types of costs are discussed. The foundation of cost-sensitive rough sets is presented. Then the current situation is summarized from the four levels including the data model, the computational model, the problem and the algorithm. Finally, research trends and key issues of test-cost-sensitive rough set are highlighted.