【摘要】 ID3算法是数据挖掘中经典的分类算法。它往往选择取值较多的属性进行分裂训练集,而选取的这个属性并非是最优的。针对这一缺点,提出了一种改进的ID3算法。通过改进信息增益公式,选取最优的划分属性,对采集的数据进行分类处理,建立决策树,这样的决策树包括较少的分支,并且树的高度较低。改进后的算法结合模式匹配算法来检测是否有入侵行为发生。通过实验验证了该算法减少了误报率和漏报率,并且比修改前在速度上有所提高,空间消耗有所减少。更多还原

【Abstract】 ID3 algorithm is a classic classification of data mining.It always selected the attribute,which has many values,but it wasn’t the best one.To this fault,an improved decision tree algorithm was proposed.Though improved information gain,selecting the best division attributes,the data collected was classified,then the decision tree with less branches and lower height was created.The algorithm,combining with pattern match,was used to detect intrusions.Experimental results showed the effectiveness of the algorithm,false alarm rate and omission rate decreased,increasing the detection rate and reducing the space consumption.更多还原