An Algorithms for Attribute Reduction in Decision Table Based on Information Capacity in Rough Set

In this paper,an improved information quantity-based heuristic algorithm for reduction of attribute is proposed.Also,it presents a concept of conditional information capacity such as decision attribute set and relative condition attribute set,and uses the condition of knowledge information content to define the importance of the property.On this foundation,we put forward a kind of new algorithm based on information content where the complexity of it is O(|C|3|U|2,which can be shown that this algorithm is effective is effective and can be analyzed by practical example.