摘要
在小波分析的基础上,运用支持向量机(SVM)方法来对噪声和非噪声数据进行分类。首先,把一带有噪声的信号进行多尺度小波分解;然后通过试验检测出小波分解系数中部分噪声信号和非噪声信号,得到样本数据来训练SVM;最后对所有的小波系数用训练后的SVM来进行分类得到非噪声信号,并且对这部分非噪声信号进行小波重构即达到了去噪的目的。
Noised and not noised signal data are identified using SVM method on the basis of wavelet transform. Firstly, noised signal data are decomposed by multilevel transform; Secondly, train some noised and not noised data of by testing; Finally, all coefficients are classed by SVM and obtains the not noised signal data. Reconstructed signal is obtained by using the inverse wavelet transform for not noised signal coefficients.
出处
《湖南科技学院学报》
2005年第5期157-159,共3页
Journal of Hunan University of Science and Engineering
基金
湖南省自然科学基金项目(03JJY3101)
湖南省教育厅科研项目(04C076)
