基于Adaboost分类算法的优化研究与应用 |
| |
引用本文: | 吴琼,周维民,李运田. 基于Adaboost分类算法的优化研究与应用[J]. 工业控制计算机, 2013, 0(12): 90-92 |
| |
作者姓名: | 吴琼 周维民 李运田 |
| |
作者单位: | 上海大学机电工程与自动化学院,上海200072 |
| |
摘 要: | 针对数据采集过程中的数据分布不平衡的问题,对非平衡数据应用数据挖掘分类算法进行分类。传统的分类器在处理非平衡数据时分类结果往往倾向于样本数目较多的类。但Adaboost算法在处理非平衡数据过程中表现出了优势,主要是对Adaboost算法进行改进和应用,采用级联的Adaboost分类器并结合SVM算法构造出分类效率更高的分类器。最后通过具体数据验证改进后算法的有效性。
|
关 键 词: | Adaboost SVM 分类 非平衡样本集 级联 分类效率 数据挖掘 |
Optimization Research and Application Based on Adaboost Classify Algorithm |
| |
Abstract: | This paper focuses on the issue of unbalanced distribution of data during data collection process.The classification algorithm of data mining is applied to classify imbalanced data.However,tradition Classifiers' results in dealing with imbalanced data also trend to the large number train sets.Adaboost algorithm has the advantage in this field.This paper mainly improves traditional Adaboost algorithm by constructing an more efficient classifier combining the cascade Adaboost classifiers with SVM algorithm.At last,specific data is used to demonstrate the effectiveness of the improved algorithm. |
| |
Keywords: | Adaboost SVM classify unbalanced train sets cascade connection classification effectiveness data mining |
本文献已被 维普 等数据库收录! |
|