A. Add 10% of the rows randomly and create another model and compare the R-Square scores of both the models.

Adding more data to the training set can affect the R-Square score, but it does not directly verify the correctness of the implementation.

B. Train various models by changing the order of input features and verify that the R-Square score of these models vary significantly.

Changing the order of input features should not significantly affect the R-Square score if the implementation is correct, but this approach is more about testing model robustness rather than correctness of the implementation.

C. Compare the R-Square score of the model obtained using two different implementations that utilize two different programming languages while using the same algorithm and the same training and testing data.

This approach directly compares the performance of two implementations of the same algorithm. If both implementations produce similar R-Square scores on the same training and testing data, it suggests that the new implementation "X" is correct.

D. Drop 10% of the rows randomly and create another model and compare the R-Square scores of both the models.

Dropping data can lead to variations in the R-Square score but does not directly verify the correctness of the implementation.

Therefore, option C is the most appropriate strategy because it directly compares the performance of the new implementation "X" with another implementation using the same algorithm and datasets, which helps in verifying the correctness of the implementation​​.

The ISTQB CT-AI syllabus explains in Section 1.6 – Machine Learning Approaches that supervised learning is appropriate when labeled data exists and the goal is to predict an output based on known historical examples. Predicting a customer’s purchase probability is a classification task when the output corresponds to discrete categories such as “likely to purchase” vs. “not likely to purchase.” The syllabus gives similar examples in describing classification as the process of assigning instances to predefined classes based on learned patterns in labeled data. Because the retail company wants to determine whether a customer will make a purchase based on marketing actions, classification is the most appropriate choice .

Option A (regression) predicts continuous numeric values and is less suitable because the task centers on categorical likelihood , not estimating exact monetary values. Option C (unsupervised learning) is used when labels are not available—here, the company has labeled purchase histories. Option D (reinforcement learning) requires an interactive environment with reward-driven behavior, which is not applicable to this scenario.

Thus, supervised learning (classification) is the most suitable approach according to the syllabus.

The ISTQB CT-AI syllabus explains in Section 4.5 – Testing AI-Specific Risks that adversarial testing is a structured test activity in which testers apply adversarial attacks —crafted or perturbed inputs—to intentionally expose weaknesses in the ML model. The purpose is to identify vulnerabilities that could be exploited through data poisoning , evasion attacks , or input manipulation . Option C correctly reflects this syllabus definition: adversarial testing is about using attacks to locate weaknesses so they can be removed or mitigated .

Option A is incorrect because regression testing does not verify data sourcing policies; it verifies unchanged functionality after modifications. Option B is incorrect because adversarial examples are often added to training datasets to improve robustness (a practice called adversarial training), not excluded. Option D is incorrect because AIB testing is not described as superior to exploratory data analysis in outlier detection; both have different purposes, and EDA remains essential for data quality assessment.

Thus, Option C is consistent with syllabus-defined adversarial testing.

The ISTQB CT-AI syllabus describes the ML data preparation workflow in Section 2.2 – Data Preparation . Data preparation consists of data gathering , cleaning , transformation , and sampling . The syllabus emphasizes that one significant challenge during data gathering is combining data from multiple heterogeneous sources , which often differ in structure, quality, and format. Ensuring the resulting dataset is accurate, complete, and representative can be complex, making this a critical challenge in the ML workflow. This aligns directly with Option C .

Option A is incorrect because erroneous data correction is part of cleaning , not transformation. Option B contradicts the syllabus: while automation can help, not all steps should be automated due to the need for expert oversight, especially in detecting subtle data quality issues. Option D is incorrect because sampling continues to involve risk—particularly around representativeness—and the syllabus emphasizes caution, not complacency.

Thus, Option C is the only statement that accurately reflects the syllabus.

A. 64-bit CPUs.

While 64-bit CPUs are essential for handling large amounts of memory and performing complex computations, they are not specifically optimized for the types of operations commonly used in machine learning.

B. Hardware supporting fast matrix multiplication.

Matrix multiplication is a fundamental operation in many machine learning algorithms, especially in neural networks and deep learning. Hardware optimized for fast matrix multiplication, such as GPUs (Graphics Processing Units), is most suitable for implementing AI and ML because it can handle the parallel processing required for these operations efficiently.

C. High powered CPUs.

High powered CPUs are beneficial for general-purpose computing tasks and some aspects of ML, but they are not as efficient as specialized hardware like GPUs for matrix multiplication and other ML-specific tasks.

D. Hardware supporting high precision floating point operations.

High precision floating point operations are important for scientific computing and some specific AI tasks, but for many ML applications, fast matrix multiplication is more critical than high precision alone.

Therefore, the correct answer is B because hardware supporting fast matrix multiplication, such as GPUs, is most suitable for the parallel processing requirements of machine learning​​.

A/B testing, also known as split testing, is a method used to compare two versions of a product or system to determine which one performs better. It is widely used in web development, marketing, and machine learning to optimize user experiences and model performance. Here’s why option C is the least descriptive of an A/B testing scenario:

Understanding A/B Testing:

In A/B testing, two versions (A and B) of a system or feature are tested against each other. The objective is to measure which version performs better based on predefined metrics such as user engagement, conversion rates, or other performance indicators.

Application in Machine Learning:

In ML systems, A/B testing might involve comparing two different models, algorithms, or system configurations on the same set of data to observe which yields better results.

Why Option C is the Least Descriptive:

Option C describes comparing the performance of an ML system on two different input datasets. This scenario focuses on the input data variation rather than the comparison of system versions or features, which is the essence of A/B testing. A/B testing typically involves a controlled experiment with two versions being tested under the same conditions, not different datasets.

Clarifying the Other Options:

A. A comparison of two different websites for the same company to observe from a user acceptance perspective : This is a classic example of A/B testing where two versions of a website are compared.

B. A comparison of two different offers in a recommendation system to decide on the more effective offer for the same users : This is another example of A/B testing in a recommendation system.

D. A comparison of the performance of two different ML implementations on the same input data : This fits the A/B testing model where two implementations are compared under the same conditions.

The ISTQB CT-AI syllabus explains in Section 3.2 – Functional Performance Criteria of ML Models that accuracy becomes unreliable when class imbalance exists. In fraud detection, more than 99% of transactions are non-fraudulent, meaning the dataset is extremely imbalanced. Because accuracy counts all correct non-fraudulent classifications, it will appear artificially high, even if the fraud detection performance is poor. Therefore, accuracy is not suitable for evaluating fraud detection systems.

The syllabus further explains that sensitivity (recall) captures the proportion of correctly identified fraudulent cases. This metric is important, as missing fraudulent events can cause high financial loss. However, the client also stresses that legitimate transactions must be correctly identified , meaning false positives must be minimized to maintain customer satisfaction.

The F1 score , defined as the harmonic mean of precision and recall , balances both:

Precision protects legitimate customers by minimizing false alarms.

Recall ensures fraudulent transactions are detected.

Section 3.2 emphasizes that when both false positives and false negatives have significant consequences, and the data is highly imbalanced, F1 is the most appropriate metric because it reflects the combined importance of detecting fraud while avoiding unnecessary alerts. Thus, Option C is the correct choice.