Bond Credit Testing System and Methods

Abstract

A bond credit detection system includes a storage unit and a processing unit. The storage unit stores a financial indicator extraction module, an optimized prediction probability module, and a neural network prediction module based on a learning rate accelerator. The processing unit is electrically connected to the storage unit and loads the financial indicator extraction module. It receives a bond international security identification code to obtain multiple financial indicator data corresponding to the bond international security identification code. Then, the processing unit loads the optimized prediction probability module and filters multiple key predictive financial indicator factors from the financial indicator data according to a logarithmic probability model. Next, the processing unit loads the neural network prediction module and inputs the key predictive financial indicator factors. Finally, the neural network prediction module outputs a downgrade probability value.

Images