基于FTT和完全卷积神经网络的发症分类嵌套了LSTM

Jianhao Nie1, Huazhong Shu1, Fuzhi Wu1

  • 1Laboratory of Image Science and Technology, Key Laboratory of Computer Network and Information Integration, Ministry of Education, Southeast University, Nanjing, China.

PubMed
概括

本研究介绍了一种先进的症分类方法,使用快速里埃转换 (FFT) 功能与卷积神经网络 (CNN) 和长短期记忆 (LSTM) 模型. 该方法的准确性超过97%,有助于有效诊断.