| Topic | Details |
| Topic 1 | - ML Model Development: This section of the exam measures skills of Fraud Examiners and covers choosing and training machine learning models to solve business problems such as fraud detection. It includes selecting algorithms, using built-in or custom models, tuning parameters, and evaluating performance with standard metrics. The domain emphasizes refining models to avoid overfitting and maintaining version control to support ongoing investigations and audit trails.
|
| Topic 2 | - Data Preparation for Machine Learning (ML): This section of the exam measures skills of Forensic Data Analysts and covers collecting, storing, and preparing data for machine learning. It focuses on understanding different data formats, ingestion methods, and AWS tools used to process and transform data. Candidates are expected to clean and engineer features, ensure data integrity, and address biases or compliance issues, which are crucial for preparing high-quality datasets in fraud analysis contexts.
|
| Topic 3 | - ML Solution Monitoring, Maintenance, and Security: This section of the exam measures skills of Fraud Examiners and assesses the ability to monitor machine learning models, manage infrastructure costs, and apply security best practices. It includes setting up model performance tracking, detecting drift, and using AWS tools for logging and alerts. Candidates are also tested on configuring access controls, auditing environments, and maintaining compliance in sensitive data environments like financial fraud detection.
|
| Topic 4 | - Deployment and Orchestration of ML Workflows: This section of the exam measures skills of Forensic Data Analysts and focuses on deploying machine learning models into production environments. It covers choosing the right infrastructure, managing containers, automating scaling, and orchestrating workflows through CI
- CD pipelines. Candidates must be able to build and script environments that support consistent deployment and efficient retraining cycles in real-world fraud detection systems.
|