支持向量机在长江芜湖段水质评价中的应用

Study of water quality evaluation based on rough sets and wavelet support vector machines

为了快捷和高精度地评价水质,针对支持向量机的训练数据量局限于小样本集以及对噪音数据的敏感性问题,提出了一种基于粗糙集与Morlet小波核支持向量机的水质评价方法.利用本算法和matlab平台在长江芜湖段15项参评指标检测数据的108个样本基础上,进行水质评价建模和分类.实验表明,利用小波核不仅提高了分类的准确性,而且提高了整体分类效率.

Aiming at the problem of support vector machines that training data bulk is limited to small pattern set and it is sensitive to the noise data,water quality evaluation based on rough sets and wavelet support vector machines was advanced. This method employs the reduction nature of rough set theory as water quality evaluation preprocessor, clearing up the redundancy of sample and noise data. Then, based on the Morlet wavelet and support vector machines, modelling was setup to sort the water quality evalua- tion. Experiment indicates that using wavelt not only improvs the classification accuracy but also raises the efficiency of global classification.