Explainability in AI-based systems refers to the ease with which users can determine how the system reaches a particular result. It is a crucial aspect when demonstrating AI decision-making, as it ensures that decisions made by AI models are transparent, interpretable, and understandable by stakeholders.
Before deploying an AI-based system, a developer must validate how decisions are made in a test environment. This process falls under the characteristic of explainability because it involves clarifying how an AI model arrives at its conclusions, which helps build trust in the system and meet regulatory and ethical requirements.
* ISTQB CT-AI Syllabus (Section 2.7: Transparency, Interpretability, and Explainability)
* "Explainability is considered to be the ease with which users can determine how the AI-based system comes up with a particular result".
* "Most users are presented with AI-based systems as 'black boxes' and have little awareness of how these systems arrive at their results. This ignorance may even apply to the data scientists who built the systems. Occasionally, users may not even be aware they are interacting with an AI- based system".
* ISTQB CT-AI Syllabus (Section 8.6: Testing the Transparency, Interpretability, and Explainability of AI-based Systems)
* "Testing the explainability of AI-based systems involves verifying whether users can understand and validate AI-generated decisions. This ensures that AI systems remain accountable and do not make incomprehensible or biased decisions".
* Contrast with Other Options:
* Autonomy (B): Autonomy relates to an AI system's ability to operate independently without human oversight. While decision-making is a key function of autonomy, the focus here is on demonstrating the reasoning behind decisions, which falls under explainability rather than autonomy.
* Self-learning (C): Self-learning systems adapt based on previous data and experiences, which is different from making decisions understandable to humans.
* Non-determinism (D): AI-based systems are often probabilistic and non-deterministic, meaning they do not always produce the same output for the same input. This can make testing and validation more challenging, but it does not relate to explaining the decision-making process.
Supporting References from ISTQB Certified Tester AI Testing Study Guide:Conclusion:Since the question explicitly asks about the characteristic under which decision-making falls when being demonstrated before deployment,explainability is the correct choicebecause it ensures that AI decisions are transparent, understandable, and accountable to stakeholders.

Detecting biases in the ML pipeline involves various tests to ensure fairness and accuracy throughout the ML process.
Testing the distribution shift in the training data for inappropriate bias (A): This involves checking if there is any shift in the data distribution that could lead to bias in the model. It is an important test but not the most direct method for detecting biases.
Test the model during model evaluation for data bias (B): This is a critical stage where the model is evaluated to detect any biases in the data it was trained on. It directly addresses potential data biases in the model.
Testing the data pipeline for any sources for algorithmic bias (C): This test is crucial as it helps identify biases that may originate from the data processing and transformation stages within the pipeline. Detecting sources of algorithmic bias ensures that the model does not inherit biases from these processes.
Check the input test data for potential sample bias (D): While this is an important step, it focuses more on the input data and less on the overall data pipeline.
Hence, the most likely useful test to help detect different kinds of biases in the ML pipeline is B. Test the model during model evaluation for data bias.
Reference:
ISTQB CT-AI Syllabus Section 8.3 on Testing for Algorithmic, Sample, and Inappropriate Bias discusses various tests that can be performed to detect biases at different stages of the ML pipeline.
Sample Exam Questions document, Question #32 highlights the importance of evaluating the model for biases.

The question asks which deficiency is most likely to be discovered by the test expert given the scenario of poor real-world performance despite good isolated accuracy.
* A lack of similarity between the training and testing data (A): This is a common issue in ML where the model performs well on training data but poorly on real-world data due to a lack of representativeness in the training data. This leads to poor generalization to new, unseen data.
* The input data has not been tested for quality prior to use for testing (B): While data quality is important, this option is less likely to be the primary reason for the described issue compared to the representativeness of training data.
* A lack of focus on choosing the right functional-performance metrics (C): Proper metrics are crucial, but the issue described seems more related to the data mismatch rather than metric selection.
* A lack of focus on non-functional requirements testing (D): Non-functional requirements are important, but the scenario specifically mentions issues with detecting real cancer cases, pointing more towards data issues.
:
ISTQB CT-AI Syllabus Section 4.2 on Training, Validation, and Test Datasets emphasizes the importance of using representative datasets to ensure the model generalizes well to real-world data.
Sample Exam Questions document, Question #40 addresses issues related to data representativeness and model generalization.

The syllabus describes automation bias as:
"A type of bias caused by a person favoring the recommendations of an automated decision-making system over other sources." This is also known as complacency bias, where testers accept automated system outputs without questioning them.
(Reference: ISTQB CT-AI Syllabus v1.0, Glossary, Page 89 of 99)

Decision trees are a widely usedmachine learning (ML) techniquethat falls undersupervised learning. They are used for bothclassification and regressiontasks and are popular due to their interpretability and effectiveness.
* How Decision Trees Work:
* The model splits the dataset into branches based on feature conditions.
* It continues to divide the data until each subset belongs to a single category (classification) or predicts a continuous value (regression).
* The final result is a tree structure where decisions are made atnodes, and predictions are given at leaf nodes.
* Common Applications of Decision Trees:
* Fraud detection
* Medical diagnosis
* Customer segmentation
* Recommendation systems
* B (Equivalence Partitioning):This is asoftware testing technique, not a machine learning method. It is used to divide input data into partitions to reduce test cases while maintaining coverage.
* C (Boundary Value Analysis):Anothersoftware testing technique, used to check edge cases around input boundaries.
* D (Decision Tables):A structuredtesting techniqueused to validate business rules and logic, not a machine learning method.
* ISTQB CT-AI Syllabus (Section 3.1: Forms of Machine Learning - Decision Trees)
* "Decision trees are used in classification and regression models and are fundamental ML algorithms".
Why Other Options Are Incorrect:Supporting References from ISTQB Certified Tester AI Testing Study Guide:Conclusion:Sincedecision trees are a core technique in machine learning, while the other options are software testing techniques, thecorrect answer is A.