The existing explicit learning method of associative classification can' t solve small disjunction problem and the lazy method' s classification efficiency is low. According to the deficiency of the two approaches, this paper proposed an improved algorithm, associative classification based on hybrid strategy. The algorithm could be summarized as follows. Firstly, it judged whether the test sample met the classifier characters of the explicit learning mode, then used the explicit learning method to classify the test sample which met the classifier characters and used the Lazy method to classify the test sample which didn' t meet the classifier characters. Finally, it combined the classification results of the two types of methods to get the final classification results. The experiments compared this method with the traditional associative classification approaches. Results show that the method is more effective in terms of classification accuracy and execution efficiency.%关联分类中现有的显式学习方法无法解决small disjunction问题,而Lazy方法分类效率低.针对这两类方法存在的问题,提出了一种基于混合策略的关联分类方法.具体算法为:先判断待分类样本是否满足显式学习模式的分类器特征；然后把满足分类器特征的待分类样本用显式模式进行分类,把不满足分类器特征的待分类样本用Lazy模式来预测；最后结合两类方法的分类结果得到最终的分类结果.实验比较了该方法与传统的关联分类方法,结果表明,该方法在分类准确率和执行效率方面均达到了更好的效果.
展开▼
