Applied Fraud Detection with Python

by Alice Schwartz, Takehiro Kanegi

Reactive PublishingApplied Fraud Detection with Python is a practical, systems-level guide to building modern fraud, anomaly detection, and AML infrastructure at scale.Designed for analysts, data scientists, engineers, and financial professionals, this book goes beyond toy examples to focus on real operational constraints: noisy data, evolving fraud patterns, regulatory pressure, and the need for explainable, auditable models. You'll learn how Python is used in production environments to detect suspicious behavior across transactions, users, networks, and time.The book covers the full fraud detection lifecycle, from data ingestion and feature engineering to statistical baselines, machine learning models, and real-time monitoring systems. Emphasis is placed on anomaly detection techniques, behavioral modeling, graph-based fraud analysis, and scalable pipelines suitable for banks, fintech platforms, payment processors, and compliance teams.Rather than treating fraud detection as a single model problem, this book frames it as an adaptive system, one that must balance precision, recall, latency, and regulatory transparency. Python's ecosystem is used throughout to connect analytics, modeling, and deployment into cohesive AML and risk platforms.What you'll learn:

• Designing fraud and AML systems as end-to-end pipelines
• Statistical and machine learning approaches to anomaly detection
• Feature engineering for transactional and behavioral data
• Detecting fraud using time-series and network analysis
• Building scalable, auditable fraud detection architectures
• Managing false positives, drift, and model decay in production
• Integrating fraud analytics into compliance and risk workflows

• Fraud and AML analysts
• Data scientists and machine learning engineers
• Financial engineers and risk professionals
• Developers building transaction monitoring systems
• Anyone designing large-scale trust, risk, or compliance platforms