%0 Journal Article
%T 使用机器学习算法基于行为决定因素的宫颈癌早期检测<br>Early Detection of Cervical Cancer Based on Behavioral Determinants Using Machine Learning Algorithms
%A 王文盈
%J Statistics and Applications
%P 1336-1345
%@ 2325-226X
%D 2023
%I Hans Publishing
%R 10.12677/SA.2023.125137
%X 宫颈癌是最常见的女性生殖道恶性肿瘤，发病率在女性恶性肿瘤中居第二位，在某些发展中国家甚至位居首位。在我国，宫颈癌也是危害女性健康与生命的重要疾病。幸运的是，女性可以通过接种HIV疫苗和定期筛查来预防这种疾病，但目前预防方法的结果和参与度都比较低。一方面是由于经济条件的限制，另一方面是因为女性本身预防知识和意识的缺乏。因此，为了积极预防宫颈癌，本文分别利用逻辑斯蒂回归、线性判别分析、二次判别分析、朴素贝叶斯、K-近邻方法检测基于行为决定因素的宫颈癌风险，并选择准确率最高的预测方法。而在正式进行预测分类前，本文利用主成分分析对数据进行降维处理；利用交叉验证对变量进行选择。<br />
Cervical cancer is the most common malignancy of the female reproductive tract, ranking second in incidence and even first in some developing countries. In China, cervical cancer is also an important disease that endangers women’s health and life. Fortunately, women can prevent the disease through HIV vaccination and regular screening, but current prevention methods have low outcomes and participation. On the one hand, it is due to the restrictions of economic conditions, and on the other hand, it is because of the lack of prevention knowledge and awareness of women themselves. Therefore, in order to actively prevent cervical cancer, this paper used logistic regression, linear discriminant analysis, quadratic discriminant analysis, naive Bayes and K-nearest neighbor methods to detect the risk of cervical cancer based on behavioral determinants, and selected the prediction method with the highest accuracy. Before the formal prediction classification, this paper uses principal component analysis to reduce the dimension of the data. Use cross-validation to select variables.
%K 逻辑斯蒂回归，线性判别分析，二次判别分析，朴素贝叶斯，K-近邻，主成分分析，交叉验证<br>Logistic Regression
%K Linear Discriminant Analysis
%K Quadratic Discriminant Analysis
%K Naive Bayes
%K K-Nearest Neighbor
%K Principal Component Analysis
%K Cross Verification
%U http://www.hanspub.org/journal/PaperInformation.aspx?PaperID=74165