A Deep Learning-Based Model for Financial Crisis Prediction

Deep learning (DL) is becoming more widespread in the banking and finance industries. There is currently no thorough analysis of deep learning's uses in banking and finance in the corpus of extant literature. This study examines the literature on the application of model evaluation, input data evaluation, and model preprocessing evaluation. This study gives scholars and practitioner’s guidance and insight into the current status of deep learning model application in the banking and finance industries. Two recurrent neural network architectures are compared in this study: the recurrent neural network (RNN) and the short-term memory (LSTM), the results show that deep learning is superior for financial crisis prediction. RNN has a 99.2% precision rate.