监督式分组套索法在急性白血病分型中的应用  

Application of Supervised Group Lasso in the Classification of Acute Leukemia

作　　者：苏纯燕[1] 房云[1] 

机构地区：[1]上海师范大学数理学院,上海200234

出　　处：《数理统计与管理》2016年第1期32-38,共7页Journal of Applied Statistics and Management

基　　金：国家自然科学基金(11201306);上海市教育委员会科研创新项目(13YZ065);上海师范大学项目(SK201207;DZL121);上海高校青年教师培养资助计划资助

摘　　要：急性白血病可分为急性淋巴细胞白血病(ALL)和急性髓系白血病(AML)两大亚型,准确诊断是治疗急性白血病的前提和关键。本文基于急性白血病的基因芯片数据,结合两样本T检验、Wilconxon秩和检验、系统聚类法以及变量选择方法监督式分组套索法(supervised group lasso,SGLasso)筛选出对急性白血病分型(AML、ALL)有显著意义的基因,根据训练组数据建立关于急性白血病分型的逻辑回归模型,并对训练组和检验组中患者的病型作拟合和预测,验证该模型的预测精度。Acute Leukemia can be classified into two subtypes, acute lymphoblastic leukemia （ALL） and acute myeloid leukemia （AML）. Accurate diagnosis is the premise and key for treatment. Based on the microarray data of acute leukemia, this paper chooses the genes which have significant effect on the classification of acute leukemia with the combination of two-sample T test, Wilcoxon rank sum test, hierarchical clustering and supervised group lasso （SGLasso） for variable selection, builds the logistic regression model based on the training dataset, fits and predicts the subtypes of acute leukemia for the training and testing datasets to verify the prediction performance of our model.

关 键 词：系统聚类法 逻辑回归 套索法 监督式分组套索法 极大似然估计 

分 类 号：O212[理学—概率论与数理统计]