Explainable AI for Fraud Detection: An Attention-Based Ensemble of CNNs, GNNs, and A Confidence-Driven Gating Mechanism

TL;DR

This paper tackles credit card fraud detection by proposing an attention-based stacking framework that fuses CNN, RNN, LSTM, and GNN outputs using DOWA and IOWA aggregation, guided by a confidence-aware gating mechanism. SHAP-based feature selection identifies the top 10 discriminative features to boost interpretability, while an MLP meta-learner combines the most reliable fused outputs. The method demonstrates strong performance across balanced and imbalanced datasets, with an emphasis on uncertainty handling and transparency, enabling trustworthy deployment in real-world systems. The work offers practical insights for robust, scalable fraud detection and highlights avenues for future enhancements, including online learning and alternative aggregation operators.

Abstract

The rapid expansion of e-commerce and the widespread use of credit cards in online purchases and financial transactions have significantly heightened the importance of promptly and accurately detecting credit card fraud (CCF). Not only do fraudulent activities in financial transactions lead to substantial monetary losses for banks and financial institutions, but they also undermine user trust in digital services. This study presents a new stacking-based approach for CCF detection by adding two extra layers to the usual classification process: an attention layer and a confidence-based combination layer. In the attention layer, we combine soft outputs from a convolutional neural network (CNN) and a recurrent neural network (RNN) using the dependent ordered weighted averaging (DOWA) operator, and from a graph neural network (GNN) and a long short-term memory (LSTM) network using the induced ordered weighted averaging (IOWA) operator. These weighted outputs capture different predictive signals, increasing the model's accuracy. Next, in the confidence-based layer, we select whichever aggregate (DOWA or IOWA) shows lower uncertainty to feed into a meta-learner. To make the model more explainable, we use shapley additive explanations (SHAP) to identify the top ten most important features for distinguishing between fraud and normal transactions. These features are then used in our attention-based model. Experiments on three datasets show that our method achieves high accuracy and robust generalization, making it effective for CCF detection.

Figures (12)

• Figure 1: SHAP-based feature importance ranking for fraud detection.
• Figure 2: Traditional stacking classifier
• Figure 3: The block diagram of proposed architecture for detecting CCF.
• Figure 4: Correlation matrix illustrating the diversity among predictions from RNN, GNN, CNN, and LSTM. Darker shades indicate higher agreement, whereas lighter shades reflect lower correlation.
• Figure 5: Confusion matrix for Dataset 1, using the proposed ensemble model with an 80-20 stratified train-test split.