The Pre-warning Of Financial Crisis In China--Based On The Model Of Artificial Neural Network

The paper analyses feasibility of artificial neural network (ANN) pre-warning financial crisis, and the model of artificial neural network (ANN) is trained by economic data from 1991 to 2001 and tested by economic data of 2002 within certain error, then ,the chained model will be analyzed economic data of 2003 and pre-warns financial crisis of 2004. finally, by the model, we deduced that pre-warning result of financial risk in 2004 is level III, and mostly come from macro-economic risk, serious risk of finance and bank.The structure of the paper follows as: firstly, briefly introduces and comments on primary methods of pre-warning financial crisis, analyses the feasibility of artificial neural network (ANN) pre-warning financial crisis, secondly, the model of artificial neural network (ANN) is trained by economic data from 1991 to 2001, and tested by economic data of 2002 within certain error, then, the chained model will be analyzed economic data of 2003 and be pre-warned financial crisis of 2004, analyses pre-warning result, lastly, the paper give some countermeasure and advice.