فهرست نشریات علمی دسترسی آزاد ایران

Customer credit risk assessment is crucial for the financial stability of leasing companies. While Data Envelopment Analysis/Discriminant Analysis (DEA/DA) models have been widely used in finance, limited attention has been given to their application under fuzzy data environments in the leasing sector. This study proposes an integrated fuzzy DEA/DA framework designed to classify and predict customer creditworthiness when financial data are imprecise or uncertain. The framework combines the efficiency evaluation of DEA with the classification capability of DA within a fuzzy optimization structure. It employs the 6C framework and a Delphi-based expert consensus process to identify eight key indicators relevant to leasing credit risk. The model is tested on a real-world dataset of 83 customers from a leasing company, effectively classifying customers into creditworthy and non-creditworthy groups and estimating their membership degrees. Comparative analysis with several  DA techniques demonstrates that the proposed model achieves higher classification accuracy and stronger robustness to data uncertainty. Additionally, a comparative evaluation with Support Vector Machine (SVM) and Random Forest (RF) classifiers was conducted, confirming that all models perform strongly. However, the fuzzy DEA/DA model provides a superior balance between predictive accuracy and interpretability, making it a transparent and reliable tool for credit-risk assessment. Rather than introducing a new methodological form of DEA/DA, the study’s contribution lies in contextual adaptation and empirical validation of the fuzzy DEA/DA approach for the leasing industry, offering a transparent and practical decision-support tool for managing credit risk under uncertainty.