基因表达数据小波降噪与肿瘤识别

建立了基于小波降噪和支持向量机的结肠癌基因表达数据肿瘤识别模型.对试验数据进行小波分解,并利用交叉验证的方法计算试验样本的平均分类准确率,确定小波函数与小波分解层数;引入能量阈值方法对小波分解系数进行阈值处理,达到降噪的目的;提出了基因分类贡献率与主成分分析结合的方法,提取结肠癌样本数据特征;利用支持向量机强大的非线性映射能力,实现对结肠癌样本数据的非线性分类.为了减弱样本集的划分对分类准确率的影响,本文采取Jackknife检验方法对支持向量分类器的分类器检验,其分类准确率为96.77%.试验结果证明了该方法的有效性,该方法对结肠癌的识别具有一定的参考价值.

Wavelet De-Noising and Identification of Tumor of Gene Expression Dada

Recognition model of colon cancer tumor Gene Expression Data was established by using wavelet noise reduction and support vector machine（SVM）.The wavelet decomposition was made based on the test data and the method of Cross-validation was used to calculate the average classification accuracy rate of test sample in order to determine the wavelet function and wavelet decomposition level;The Energy threshold method was introduced to process wavelet coefficients for achieving noise reduction purposes;The combination method of contribution rate of gene classification and Principal component analysis（PCA） was proposed to extract characteristics of colon cancer sample data;According to the powerful nonlinear mapping ability of support vector machine,colon cancer sample data was nonlinear classified.For weakening the impact of dividing sample set on classification accuracy,Prediction accuracy of the sample set was calculated by jackknife test.The classification accuracy rate is 96.77%.Experimental results show the effectiveness of the method.The research method in this paper will be valuable to research the identification of colon cancer.