基于粗糙集理论的示例学习研究

Study on learning from examples based on rough sets theory.

到目前为止 ,一些启发式算法被提出用于基于扩张矩阵理论的示例学习研究 .该文基于粗集理论研究示例学习问题 ,提出了粗集理论下的几个新概念 ,如 :必要选择子 ,核选择子集 ,约简选择子集和所产生复合的评价指标 :精确度、覆盖度和简单性 ,给出了扩张矩阵的粗糙集算法 ,并提出了基于覆盖度和简单性的遗传算法最优示例学习方法 .

Up to now, some heuristic algorithms have been proposed for learning from examples based on extension matrix theory. The approach of learning from examples in this paper is based on rough set theory. Several new concepts are proposed in this paper, such as indispensable selector, core set of selectors, reducted set of selectors as well as accuracy, coverage and simplicity for evaluating a complex. The algorithms for resolving extension matrix and a GA method of optimal learning from examples are also represented.