A Batch Normalization Autoencoder Model for Breast Cancer Multidimensional Follow-up Data
Xuan Liu, Zhiguo Shi, Zhang Xue, Yang Chun · 2018
The traditional follow-up model is insufficient in terms of processing speed and accuracy, which results in poor analysis. In addition, a unidimensional follow-up data limits the reliability of analysis. Aimed at the problems, a deep neural network model is constructed, that is based on stacked denoising sparse autoencoder with batch normalization algorithm to deal with data of ten dimensions. It not only improves the learning ability of the autoencoder, but also enhances the prediction ability of the follow-up model. The follow-up data sets are derived from Peking University First Hospital, the Second Hospital of Shandong University and so on. ROC curve and AUC are used as evaluation indicators. The experiments indicate the model has advanced performance in both accuracy and stability for prediction than traditional machine learning algorithms.