Unlocking the Black Box: A Five-Dimensional Framework for Evaluating Explainable AI in Credit Risk

TL;DR

This paper introduces a five-dimensional framework for evaluating explainability in credit risk models, enabling the use of complex machine learning models while satisfying regulatory transparency requirements.

Contribution

It proposes a novel five-dimensional framework for assessing model explainability, bridging the gap between high-performance models and regulatory interpretability needs.

Findings

Complex models can achieve comparable explainability with LIME and SHAP.

The framework facilitates balanced trade-offs between accuracy and interpretability.

High-performing models are feasible in regulated environments using the proposed evaluation.

Abstract

The financial industry faces a significant challenge modeling and risk portfolios: balancing the predictability of advanced machine learning models, neural network models, and explainability required by regulatory entities (such as Office of the Comptroller of the Currency, Consumer Financial Protection Bureau). This paper intends to fill the gap in the application between these "black box" models and explainability frameworks, such as LIME and SHAP. Authors elaborate on the application of these frameworks on different models and demonstrates the more complex models with better prediction powers could be applied and reach the same level of the explainability, using SHAP and LIME. Beyond the comparison and discussion of performances, this paper proposes a novel five dimensional framework evaluating Inherent Interpretability, Global Explanations, Local Explanations, Consistency, and Complexity to offer a nuanced method for assessing and comparing model explainability beyond simple accuracy metrics. This research demonstrates the feasibility of employing sophisticated, high performing ML models in regulated financial environments by utilizing modern explainability techniques and provides a structured approach to evaluate the crucial trade offs between model performance and interpretability.