不平衡数据分类及在疾病诊断中的应用研究进展

Research progress of unbalanced data classification and its application in disease diagnosis

不同类之间数据分布的不平衡称为类间不平衡,不平衡数据中不同类别的样本数量差异很大。近年,机器学习诊断成为疾病诊断、分析和预测的流行选择,他比传统诊断方法省时,且能够直接预测引起疾病的相关因素,以减轻危害,减少个人和社会的负担。本文综述了近年常用的不平衡数据处理的机器学习方法,重点分析其在预测恶性肿瘤、心脏病、糖尿病及其并发症和其他疾病中的应用情况。

The imbalance of data distribution between different classes is called the imbalance between classes.The number of samples in different categories in unbalanced data varies greatly.In recent years,machine learning algorithm has been a popular choice for diagnosis,analysis and prediction of diseases,because it saves time than traditional diagnosis methods,and can directly predict the relevant factors causing diseases to reduce the harm and reduce the burden of individuals and society.The paper reviews the machine learning methods used in imbalanced data processing in recent years,focusing on their application in predicting cancer,heart disease,diabetes and its complications,and other diseases.