Balancing Performance and Reject Inclusion: A Novel Confident Inlier Extrapolation Framework for Credit Scoring

2025-10-14

AI Summary

In credit scoring, sample bias arises from application rejections—a challenge known as rejection inference (RI)—which hampers model generalization to the full applicant population. Method: This paper proposes Confidence-Inlier Extrapolation (CI-EX), a novel framework that jointly leverages anomaly detection and probabilistic supervised classification. CI-EX employs a distribution-aware confidence-inlier selection mechanism to iteratively identify and reliably label high-confidence “inliers” among rejected applicants, avoiding the blind extrapolation inherent in conventional RI methods. Contribution/Results: To better evaluate RI performance, we introduce Area under the Kickout (AuK), a complementary metric to AUC that balances discriminative power with inference inclusivity. Extensive experiments on two large-scale real-world credit datasets demonstrate that CI-EX consistently outperforms state-of-the-art baselines on RI-specific metrics while maintaining competitive AUC. The results substantiate CI-EX’s enhanced effectiveness and robustness for practical rejection inference.

Technology Category

• Machine Learning: Calibration & Uncertainty Quantification
• Cognitive Modeling & Cognitive Systems: Conceptual Inference and Reasoning
• Reasoning under Uncertainty: Other Foundations of Reasoning under Uncertainty

Application Category

• User Modeling, Personalization and Recommendation: Fairness-aware retrieval and ranking
• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking
• Economics, Online Markets and Human Computation: Trust and reliance of crowd workers and data experts on GenAI

Abstract

Reject Inference (RI) methods aim to address sample bias by inferring missing repayment data for rejected credit applicants. Traditional approaches often assume that the behavior of rejected clients can be extrapolated from accepted clients, despite potential distributional differences between the two populations. To mitigate this blind extrapolation, we propose a novel Confident Inlier Extrapolation framework (CI-EX). CI-EX iteratively identifies the distribution of rejected client samples using an outlier detection model and assigns labels to rejected individuals closest to the distribution of the accepted population based on probabilities derived from a supervised classification model. The effectiveness of our proposed framework is validated through experiments on two large real-world credit datasets. Performance is evaluated using the Area Under the Curve (AUC) as well as RI-specific metrics such as Kickout and a novel metric introduced in this work, denoted as Area under the Kickout. Our findings reveal that RI methods, including the proposed framework, generally involve a trade-off between AUC and RI-specific metrics. However, the proposed CI-EX framework consistently outperforms existing RI models from the credit literature in terms of RI-specific metrics while maintaining competitive performance in AUC across most experiments.

Problem

• Infer missing repayment data for rejected credit applicants
• Mitigate sample bias in credit scoring through reject inference
• Balance performance metrics with reject-specific evaluation criteria

Innovation

• Iteratively identifies rejected client distribution using outlier detection
• Assigns labels based on probabilities from supervised classification model
• Outperforms existing models on reject inference metrics while maintaining AUC