Course Summary
ISTQB: Certified Tester AI Testing (CT-AI) is designed to provide you with the knowledge and practical skills to test systems that contain artificial intelligence and machine learning – and to do so in a structured, risk-based way. The course follows the ISTQB CT-AI syllabus v2.0 and covers the specifics of AI-based systems, their quality characteristics, the machine learning workflow, data and model testing, and the two additional test levels that ML systems require. You will work hands-on with ML models, performance metrics and test techniques such as metamorphic, back-to-back and adversarial testing, so you can contribute to an ML test strategy and design test cases for real AI systems.
What you will learn:
• You will understand the current state of AI, including generative AI, and how AI-based systems differ from conventional software.
• You will gain experience implementing and testing machine learning models, including simple neural networks.
• You will learn to assess AI quality characteristics such as bias, robustness, transparency and explainability (ISO/IEC 25059).
• You will learn to calculate and interpret ML functional performance metrics.
• You will understand the ML-specific test levels and how to contribute to an ML test strategy.
• You will learn to design and execute test cases for input data and ML models.
Module 1: Introduction to AI
• AI, narrow/general/super AI and AI-based versus conventional systems
• AI technologies, generative AI, hardware and hosting
• ML frameworks, regulations and standards
Module 2: Quality Characteristics for AI-Based Systems
• ISO/IEC 25059 quality model for AI
• Bias, transparency, explainability, robustness and acceptance criteria
Module 3: Machine Learning
• Forms of machine learning and the ML workflow
• ML data: preparation, labelling and quality
• Performance metrics and neural networks (hands-on)
Module 4: Testing AI-Based Systems
• ML-specific test levels
• Test strategy and risk-based testing for AI
Module 5: Input Data Testing for ML Systems
• Bias, representativeness and labels
• Data pipelines and data quality
Module 6: Model Testing for ML Systems
• Adversarial, metamorphic and back-to-back testing
• A/B testing, experience-based testing and model drift
Module 7: ML Development Testing
• Testing in the ML development process
• Exam preparation
Other Popular Courses
CompTIA: SecurityX
- Duration: 5 Days
- Language: English
- Level: Advanced
- Exam: CAS-005
Executive Cyber Risk Certification (ECRC)
- Duration: 2 Days
- Language: English
- Level: Advanced
- Exam: ECRC
Mastering Communication & Presentation Te...
- Duration: 5 Days
- Language: Danish
- Level: Intermediate
- Exam: MCPT