【摘要】 利用最小二乘支持向量机和文献[10]中的半监督学习算法,我们对鼻咽癌患者5年生存状态进行了预测。实验结果表明:当已标注数据比较少时,两种方法的判别精度都比较低;随着已标注数据的增多,最小二乘支持向量机的推广能力逐渐增加,而半监督学习算法并没有给出更好的结果。这说明:对于鼻咽癌患者5年生存状态预测问题,最小二乘支持向量机比半监督学习方法更具有优势。更多还原

【Abstract】 In this paper,the problem of 5-year survival status prediction of patients with nasopharyngeal cancer is studied using least square support vector machine and semi-supervised learning algorithm in Ref.[10].The results show that their decision accuracies are lower when the number of labeled data is small.With increasing the labeled data,the generalized performance of least square support vector machine is improved.However,semi-supervised learning algorithm has not obtained the better results. From this, one can see that least square support vector is better than semi-supervised learning algorithm for the problem of 5-year survival status prediction of patients with nasopharyngeal cancer.更多还原