基于正则化多项式回归的癌症患者免疫检查点 阻断响应预测

浏览次数:13
  • 分享到:

摘要:

探究特征之间的非线性相互作用关系 , 构建了用于预测癌症患者免疫检查点阻断响应的正则化多项式逻辑斯蒂回归模型 . 无进展生存期和总生存期的Kaplan-Meier曲线被用来筛选单个特征和联合特征 ,并据此构建了二次多项式判别函数 . 结合多项式负对数损失函数和弹性网络惩罚函数 ,提出了一种正则化多项式逻辑斯蒂 回归模型 ,并通过特征拓维将其转化为线性模型求解 . 在泛癌 、黑色素瘤 、非小细胞肺癌和其他癌症数据集上与其他 6种方法进行比较 ,结果表明所提方法取得了更高的免疫检查点阻断响应精度 、F1分数和 AUC值 .

This paperexplored the nonlinearinteraction between featuresand constructed a regularized polynomiallogis- tic regression modelfor predicting immune checkpointblockade response in cancer patients. The Kaplan-Meier curves for pro- gression free survivaland overallsurvivalwereused to screen forindividualand combined features, a quadraticpolynomialdis- criminantfunction was constructed. Combining the polynomialnegativelogarithmiclossfunction and the elasticnetwork penal- ty function, a regularized polynomiallogistic regression model is proposed, and itis transformed into a linear model through feature extension. Compared with other six methods forthe datasetsofpan cancer, Melanoma, non-smallcelllung cancerand other cancers, the proposed method has achieved higher blocking response accuracy, F1   score and AUC value of immune checkpoints.

作者:

王小玉,奚晨曦 ,李钧涛

Wang Xiaoyu, XiChenxi , LiJuntao

机构地区:

郑州工商学院基础教学部;betway官方app 数学与信息科学学院 

引用本文:

王小玉,奚晨曦,李钧涛.基于正则化多项式回归的癌症患者免疫检查点阻断响应预测[J].betway官方app 学报 (自然科学版),2024,52(4):94-100.

WangXiaoyu,XiChenxi,LiJuntao. Prediction ofimmunecheckpointblock- ade response ofcancerpatientsbased on regularized polynomialregression[J] .JournalofHenan NormalUniver- sity(NaturalScience Edition) ,2024,52(4):94-100.DOI:10.16366/j.cnki.1000-2367.2023.07.31.0001.

基金:

国家自然科学基金;河南省科技攻关计划

关键词:

癌症 ;正则化 ;多项式回归 ;免疫检查点阻断

cancer; regularization; polynomialregression; immune checkpointblockade

分类号:

O224


基于改进 Resnet18网络的火灾图像识别.pdf

Baidu
map