The lending risk predicting of the folk informal financial organization from big data using the deep learning hybrid model

This article is first to predict and earlier warning folk lending risk used deep learning hybrid model, we find that the LSTM hybrid model has a higher predict accuracy on lending risk fore-casting and earlier warning of the FIFO, with an obviously improvement of the average value of forecasting accuracy. The predict accuracy of LSTM-GRU and LSTM-CNN models on lending risk forecasting of the FIFO is higher than others during COVID-19 pandemic. Therefore, we believe that the LSTM hybrid model, especially the LSTM-GRU model can better predict and early warn lending risk of the FIFO on big data.

Automated summary

This article utilizes a deep learning hybrid model to predict and provide earlier warnings for folk lending risk. The study finds that the LSTM hybrid model shows higher predict accuracy and significantly improves the average value of forecasting accuracy in FIFO lending risk forecasting and earlier warning. The LSTM-GRU and LSTM-CNN models exhibit higher predict accuracy in lending risk forecasting of the FIFO during the COVID-19 pandemic. Therefore, it is believed that the LSTM hybrid model, particularly the LSTM-GRU model, can better predict and early warn lending risk of the FIFO based on big data.