代价敏感特征选择的肿瘤分类应用

发布时间：2018-04-20 13:07

本文选题：代价敏学习 + 特征选择　；参考：《计算机工程与设计》2017年05期

【摘要】：针对肿瘤基因数据高维、小样本的特点,引入误分类代价,基于混沌遗传算法提出一种代价敏感特征选择算法。结合误分类代价矩阵与分类精度,构建基于最近邻的代价敏感特征选择适应度函数,在降低特征选择总成本的同时权衡代价成本与分类精度,采用混沌遗传算法提高搜索效率,提升算法性能。将该方法应用于肿瘤基因数据进行有效性验证,实验结果表明,该算法降低了特征空间维数,有效提高了肿瘤的分类性能。
[Abstract]:According to the characteristics of high dimension and small sample of tumor gene data, a cost of misclassification is introduced, and a cost sensitive feature selection algorithm based on chaos genetic algorithm is proposed. Combining the cost matrix of false classification and classification accuracy, the fitness function of cost sensitive feature selection based on nearest neighbor is constructed, which can reduce the total cost of feature selection while balancing the cost cost and classification accuracy. Chaos genetic algorithm is used to improve the search efficiency. Improved algorithm performance. The experimental results show that the algorithm reduces the dimension of feature space and improves the classification performance of tumor effectively.
【作者单位】：太原理工大学计算机科学与技术学院;山西职工医学院网络信息中心;中国农业银行山西省分行科技与产品管理部;
【基金】：山西省卫生厅科研基金项目(201301006) 山西省研究生优秀创新基金项目(20123030)
【分类号】：R730.2;TP18

【相似文献】