Financial Distress Predication based on Deep Learning Techniques: A survey
Deep learning (DL) is changing the way we analyze financial problems. By leveraging DL, we can build models that are significantly more efficient and accurate at identifying signs of financial distress. This work explores recent advancements in the use of DL for financial forecasting, especially on text data sourced from reports, websites, and news articles. It fills the gap in existing literature by providing a critical examination of the DL techniques in analyzing financial text to address the Financial Distress Prediction (FDP) problem, and provides key recommendations for future research. It reviews existing studies, discusses applications and limitations, and outlines future directions and challenges based on articles published between 2020 and 2025.