基于最小二乘模糊支持向量机的基因分类研究*

随着大量基因表达数据的涌现，把海量的数据划分成数量相对较少的组，有助于提取对生理学和医药学等有价值的生物信息。基因分类技术能够很好地处理和分析这些基因数据。提出了一种应用于基因分类的模糊最小二乘支持向量机方法，通过设置模糊隶属度改变分类中样本的贡献属性。该方法不仅考虑了样本与类中心点的距离关系，还充分考虑样本与样本之间的关系，减弱噪声或野值样本对分类的影响。采用美国威斯康星乳腺癌数据和皮马印第安人糖尿病数据进行实验检测，均取得了很好的效果。

Classification of genes based on least squares fuzzy support vector machines

With the emergence of a large amount of genetic data, divided the mass of data into a relatively small number of group can help to extract of physiology and medicine and other valuable biological information. Gene classification techniques can be very good at handling and analyzing the genetic data. This paper proposed fuzzy least squares support vector machine method which applied to gene classification. Defined the contribution of each sample by setting the fuzzy memberships.By considering the distance not only between the types of samples and the center of classification, but also between the samples and samples, noises and outliers were removed. The performance of the proposed method on the United States Wisconsin breast cancer database (WDBC) data and Pima Indians diabetes(PID) data are all achieved very good results.