小波去噪在分类问题上的应用

分类是数据挖掘领域中一项非常重要的任务,通过含噪数据求出有效决策函数十分重要.目前已有一些具体方法对含噪数据进行选择处理,从而求解决策函数.本文针对支持向量机中的二类分类问题,以一维情形为例,给出了文献1]不同的处理方法,即用小波技术对数据点集的分布函数去噪,并根据新的训练点集及对应的权值求得二类分类问题的决策函数.

The Application of Wavelet Noise Removal on Classification Problem

Classification is an important task in the field of data mining. It becomes critical to determine the decision function through data samples that suffered from noise. At present, there exist many specific methods to deal with this noise data for obtaining decision function. Based on the binary classes classification problem in support vector machines （SVMs）, we propose a wavelet noise removal method to deal with one dimension data samples, namely, noise on distribution function of data samples is removed, then the decision function of the binary classes classification problem is deduced according to the new training samples and the corresponding weights.