Page 33 - Ebook HT QTNHS
P. 33
monitoring, and non-performing loan resolution. The findings indicate that integrating traditional
and non-traditional data sources, alongside machine learning models and real-time analytics
technologies, can enable early risk detection, flexible credit behavior tracking, and more effective
bad debt resolution. Additionally, the study highlights the essential role of data infrastructure,
analytical capabilities, and regulatory sandboxes in fostering financial innovation. The proposed
model and recommendations serve as a practical reference for Vietnamese banks in their digital
transformation journey and in strengthening competitiveness within the data-driven economy.
Keywords: banking big data, digital credit risk, AI credit scoring, banking digital
transformation, machine learning models.
31
