CAJ | 학술논문

针对临床分类诊断中普遍存在的样本不均衡、错分代价不同、大量无标签样本和测量误差等特点，引入了机器学习中较新的研究成果———多层降噪自编码（stacked denoising autoencoders，SDA）神经网络，并与欠采样局部更新的元代价（metacost）算法相结合，对 SDA 神经网络进行了改进，使组合模型具有代价敏感、降低不均衡性、有效利用无标签样本、抗噪声的特性。实验中将改进的 SDA 神经网络与 SOFTMAX 回归、反向传播（back propagation，BP）神经网络、支持向量机（support vector machine，SVM）、传统多层自编码（stacked autoencoders， SAE）神经网络，以及传统 SDA 神经网络等作了比较。实验结果表明，改进的 SDA 神经网络的准确率、ROC 曲线下面积等均优于其他模型，提高了分类模型的辅助诊断性能。
침대림상분류진단중보편존재적양본불균형、착분대개불동、대량무표첨양본화측량오차등특점，인입료궤기학습중교신적연구성과———다층강조자편마（stacked denoising autoencoders，SDA）신경망락，병여흠채양국부경신적원대개（metacost）산법상결합，대 SDA 신경망락진행료개진，사조합모형구유대개민감、강저불균형성、유효이용무표첨양본、항조성적특성。실험중장개진적 SDA 신경망락여 SOFTMAX 회귀、반향전파（back propagation，BP）신경망락、지지향량궤（support vector machine，SVM）、전통다층자편마（stacked autoencoders， SAE）신경망락，이급전통 SDA 신경망락등작료비교。실험결과표명，개진적 SDA 신경망락적준학솔、ROC 곡선하면적등균우우기타모형，제고료분류모형적보조진단성능。
To aim at the common issues in clinical diagnose and classification,such as imbalance,different misclassification costs,numerous unlabeled samples,and measure errors,this paper introduced a relatively new research findings in deep lear-ning,stacked denoising autoencoders(SDA)neural network.Then SDA neural network and metacost algorithm with under-sam-pling,it combined partial update to form an improved SDA neural network.This improved SDA neural network was cost-sensi-tive,denoising,able to utilize the unlabeled samples and alleviate the imbalance.In the experiment,it compared with SOFT-MAX regression,back propagation(BP)neural network,support vector machine(SVM),stacked autoencoders(SAE)neural network,and traditional SDA neural network.It shows that the accuracy(ACC)and area under ROC(AUC)of the improved SDA neural network outperforms the others.Consequently,the auxiliary diagnostic ability of the classification model is en-hanced.