摘要
分析数据挖掘中数据预处理的重要性,提出了一种基于分类矩阵的ID3算法,通过引入分类矩阵对ID3算法的多值偏向性和分类速率进行改进,并利用实例对改进效果进行验证,最后在数据挖掘的预处理中,分析了改进算法在缺失值填充和异常数据处理中的具体应用。通过分析可以发现该改进算法能有效克服多值偏向性并提高分类速率,并在数据预处理中有很好的应用效果。
Analysis of the importance of data preproeessing in data mining, a classification algorithm is proposed based on ID3 matrix, By introducing the classification matrix, ID3 algorithm for multiple bias and classification rate is improved, and verify the improvement effect by example. Finally in the data mining preprocessing, analysis of the specific application of filling and improvement abnormal data processing algorithm in the missing value. Through the analysis can find that the improved algorithm can effectively overcome the variety bias and im prove the classification rate, and has a very good application effect in data preprocessing.
出处
《舰船电子工程》
2013年第4期28-31,共4页
Ship Electronic Engineering
关键词
数据预处理
分类矩阵
ID3
数据挖掘
data pre-processing, classification matrix, ID3, data mining
