A Data-Driven Intelligent Medical Management System via Neural Networks

Abstract

For human health, medical diagnosis plays an irreplaceable role, conventional medical diagnosis methods cannot ensure the accuracy of diagnosis due to the interference of various external factors. Therefore, this paper proposes a data-driven intelligent medical management system via neural networks(MMS-ID). The essence of this method is to predict the survival time of cancer patients with the aid of gradient boosting decision tree (GBDT) and hybrid neural network model. Firstly, GBDT screens the feature factors of matching conditions according to the set value domain, and inputs them into the neural network. Subsequently, a hybrid neural network that combines the convolutional neural network (CNN) and the long short-term memory (LSTM) model is employed to predict survival length of cancer patients. Finally, the stability of MMS-ID is analyzed and compared with a series of baseline methods. A series of experiments prove that MMS-ID has excellent performance.