Agricultural cooperatives in Thailand play a crucial role in providing deposit and credit services. Credit management faces challenges due to inadequate risk analysis, which can lead to non-performing loans (NPLs). This study develops a credit risk prediction model utilizing quantitative data, Machine Learning (ML), and Explainable AI (XAI) to improve organizational adaptation. Data from 300 cooperative members, supplemented by records from the National Statistical Office, were analyzed using the 5Cs framework (Character, Capacity, Capital, Collateral, Condition) to create profiles and predictive models employing decision trees, logistic regression, and artificial neural networks integrated with XAI. The findings indicate that Character and Collateral are most effectively modeled with decision trees, capacity with logistic regression, and Capital and Condition with neural networks combined with XAI. Accumulated local effects (ALE) highlighted key risk factors, including income, debt, default history, savings, and external variables. The integration of the 5Cs framework with ML and XAI enhances predictive accuracy, transparency, and data-driven decision-making. Change management strategies, such as personnel training, stakeholder engagement, and the promotion of technological understanding, are essential for sustainable adoption and utilization.
