Research on systemic risk early warning of China’s banking system based on recurrent neural network: theoretical framework, measurement methods and empirical analysis
A nonlinear dynamic early warning framework based on recurrent neural networks based on RNNs to enhance risk identification accuracy and timeliness is developed and demonstrated robust early warning capabilities during critical events such as the 2008 global financial crisis and the 2020 pandemic shock.