Fraud Detection Excellence—A Deep Exploration of Model Metrics, Optimization, and Governance

Abdelrahim Aqqad

Chapter Chapter 15 in Fraud Analytics in Action, 2026, pp 371-410 from Springer

Abstract: Abstract Chapter 15 advances from classification fundamentals to the metrics, optimization strategies, and governance frameworks needed to operationalize fraud detection models effectively. The chapter opens with a rigorous treatment of confusion-matrix-derived metrics—accuracy, precision, recall, specificity, F1 score, balanced accuracy, Matthews correlation coefficient, and Cohen’s kappa—before examining ROC curves and AUC as tools for threshold selection and comparative model assessment. A dedicated section explores the trade-off between model explainability and accuracy, introducing hybrid models, SHAP, and LIME as bridging techniques. The chapter then turns to optimization: learning rate dynamics, adaptive rate strategies, non-convex objective functions, batch size selection, and three hyperparameter tuning approaches—grid search, randomized search, and Bayesian optimization—each demonstrated with XGBoost code. Concepts of bias, variance, overfitting, early stopping, and regularization are developed alongside practical prevention strategies including SMOTE-based data balancing and ensemble learning. Model deployment challenges—serialization, API integration, real-time monitoring, and model drift—lead into a discussion of model governance covering audit trails, regulatory compliance, ethical fairness, and continuous monitoring. The chapter closes with AutoGluon's AutoML capabilities and two applied lab projects: predicting customer churn in telecommunications and predicting health insurance costs.

Date: 2026
References: Add references at CitEc
Citations:

There are no downloads for this item, see the EconPapers FAQ for hints about obtaining it.

Related works:
This item may be available elsewhere in EconPapers: Search for items with the same title.

Export reference: BibTeX RIS (EndNote, ProCite, RefMan) HTML/Text

Persistent link: https://EconPapers.repec.org/RePEc:spr:sprchp:978-3-032-16023-2_15

Ordering information: This item can be ordered from
http://www.springer.com/9783032160232

DOI: 10.1007/978-3-032-16023-2_15

Access Statistics for this chapter

More chapters in Springer Books from Springer
Bibliographic data for series maintained by Sonal Shukla () and Springer Nature Abstracting and Indexing ().