Free ISTQB-CT-AI Exam Dumps Preview

The ISTQB CT-AI syllabus clearly explains the structure of a deep neural network (DNN). A deep neural network consists of three types of layers: an input layer that receives the input data, one or more hidden layers composed of artificial neurons (also referred to as nodes), and an output layer that produces the final result. The defining characteristic of a deep neural network is the presence of at least one hidden layer between the input and output layers.

This directly supports Option B, which correctly states that a deep neural network has at least one internal hidden layer.

The other options are incorrect. A network with only input and output layers would not be considered deep.
There is no minimum requirement of five total layers for a network to be classified as deep. The output layer must be connected to the preceding layers in order to receive processed information; it is not isolated
to maintain integrity.

Thus, Option B is the correct statement.

The ISTQB CT-AI syllabus defines clustering as an unsupervised learning technique in which data points are grouped based on similarity, without using labeled output classes. The goal of clustering is to discover inherent structures or patterns in the data.

Option C correctly reflects this definition, stating that clustering is performed without prior knowledge of output classes.

The other options are incorrect. Classification of continuous quantities refers to regression, not clustering.
Clustering is not supervised learning, and it does not require prior knowledge of classes.

Therefore, Option C is the correct answer.

To calculate accuracy, recall, and specificity from the confusion matrix, the following values are used:

Confusion Matrix

Actually Rotten:

True Positive (TP) = 45

False Positive (FP) = 8

Actually Fresh:

False Negative (FN) = 5

True Negative (TN) = 42

Accuracy

Accuracy is the proportion of all correct predictions.

Formula:
Accuracy = (TP + TN) / (TP + TN + FP + FN)

Calculation:
Accuracy = (45 + 42) / (45 + 42 + 8 + 5)
Accuracy = 87 / 100
Accuracy = 0.87

Recall (Sensitivity)

Recall is the proportion of actual positives correctly identified.
Formula:
Recall = TP / (TP + FN)

Calculation:
Recall = 45 / (45 + 5)
Recall = 45 / 50
Recall = 0.9

Specificity

Specificity is the proportion of actual negatives correctly identified.

Formula:
Specificity = TN / (TN + FP)

Calculation:
Specificity = 42 / (42 + 8)
Specificity = 42 / 50
Specificity = 0.84

Therefore, the correct combination of accuracy, recall, and specificity is:

0.87, 0.9, 0.84

Reference ISTQB CT-AI Syllabus v1.0, Section 5.1 - Confusion Matrix and ML Functional Performance Metrics.

The ISTQB CT-AI syllabus defines AI-based autonomous systems as systems that can perceive and respond to their environment and make decisions or take actions without direct human intervention.
Autonomy implies that the system can adapt its behavior based on environmental inputs.

Option D correctly describes such a system, as the ability to respond to the environment independently is a core characteristic of AI-based autonomous functionality.

Option A is incorrect because decision-making is entirely performed by humans.
Option B is incorrect because a system that uses no software cannot implement AI-based autonomy.
Option C is incorrect because Selenium is a test automation tool and does not represent an autonomous AI system.

Therefore, Option D is the correct answer.

The process of developing a machine learning model typically involves the use of three types of datasets:

Training data
Used to train the model by learning patterns and relationships in the data.

Validation data
Used to tune hyperparameters and to reduce the risk of overfitting during training.

Test data
Used to evaluate the final model's performance and estimate how it will perform on previously unseen data.

Evaluation of the options:

A. Training data - validation data - test data
This option correctly includes all three datasets used during model development: training for learning, validation for tuning, and testing for final evaluation.

B. Training data - validation data
This option omits the test data, which is required to assess final model performance on unseen data.

C. Training data - test data
This option omits the validation data, which is important for tuning and preventing overfitting.

D. Validation data - test data
This option omits the training data, which is essential for learning the model.

Therefore, the correct answer is Option A, as it includes all datasets used during the process of learning and creating the model.

Reference ISTQB CT-AI Syllabus v1.0, Sections 3.2 and 4.1

The ISTQB CT-AI syllabus defines reward hacking as a situation in which an AI-based system achieves a specified objective by exploiting weaknesses or loopholes in the reward function, thereby fulfilling the reward criteria while violating the designer's original intent.

In this scenario, the robot vacuum exploits the reward structure by driving backwards. This allows it to maximize speed while avoiding bumper activation, even though this behavior is not aligned with the intended goal of safe forward navigation. The system technically optimizes the reward but does so in an unintended and undesirable way.

The other options do not apply. Error short-circuiting refers to bypassing error detection logic, while transparency and interpretability relate to understanding model behavior, not exploiting reward definitions.

Therefore, Reward hacking is the correct answer.

The ISTQB CT-AI syllabus describes multiple approaches to data labeling for supervised learning. Among these approaches, using a pre-labeled dataset requires the least time and effort because the labeling activity has already been completed. This eliminates the need for additional manual labeling, coordination with external resources, or human review of automatically generated labels.

Other approaches---internal labeling, outsourced labeling, and AI-assisted labeling---still require time and effort for labeling, verification, or quality control. Therefore, they are more resource-intensive than using data that is already labeled.

Thus, Pre-labeled dataset (Option B) is the correct answer.

The ISTQB CT-AI syllabus explains that exploratory data analysis (EDA) can be used to examine datasets in order to identify outliers, anomalies, and unusual patterns. In the context of adversarial attacks such as data poisoning, maliciously crafted inputs often differ statistically from legitimate data and may therefore appear as outliers.

By analyzing distributions, frequencies, and patterns in email metadata and content, EDA can help identify emails that deviate significantly from normal internal communication, indicating a potential attack.

Therefore, using EDA to detect outlier emails is an appropriate method for identifying this type of adversarial behavior, making Option A the correct answer.

Section 1.6.3 - Reinforcement Learning of the ISTQB CT-AI syllabus states that reinforcement learning (RL) is based on an agent interacting with an environment, performing actions, and receiving rewards or penalties. The core concept is the reward function, which guides the agent's learning process. The syllabus emphasizes that training in RL is driven by rewards, and the agent aims to maximize cumulative reward over time. Therefore, Option C correctly describes reinforcement learning: the agent's training is based on a reward function that rewards successful actions.

Option A is incorrect because reinforcement learning does not rely on labeled data; that is characteristic of
supervised learning.
Option B is incorrect because reinforcement learning fundamentally requires interaction with the environment.
Option D is incorrect because the reward function is defined externally (for example, by designers or domain experts); the agent learns a policy to maximize reward, not the
reward function itself.

Thus, Option C is the only statement consistent with reinforcement learning as defined in the syllabus.

The ISTQB CT-AI syllabus explains that several factors directly influence machine learning functional performance, including the data pipeline, the quality of data labeling, and the presence of biased data.
Problems in the data pipeline can introduce noise or inconsistencies, poor labeling quality can mislead the learning process, and biased data can distort model outcomes and reduce accuracy or fairness.

The number of classes, however, is primarily a characteristic of the problem formulation and model design.
While it may influence model complexity or architecture, it is not considered a direct factor affecting functional performance metrics such as accuracy, precision, recall, or bias in the sense described by the syllabus.

Therefore, Option D is the correct answer.

Using a pixel-by-pixel comparison of the GUI before and after a change is a traditional image comparison technique and does not involve AI. It performs a direct comparison of pixel values without applying learning, inference, or intelligent interpretation. Therefore, it is the least likely usage of AI for detecting GUI changes.

Using computer vision techniques to compare GUIs involves AI-based image understanding and is commonly applied to detect visual changes beyond simple pixel differences.

Using vision-based detection of GUI layout changes leverages AI to analyze structural and spatial relationships within the interface, which is a typical AI-based approach.

Using an ML-based classifier to determine whether detected GUI changes should be flagged for human review is also a valid AI usage, as it applies learned patterns to support decision-making.

Therefore, Option A correctly describes the least likely usage of AI for detecting GUI changes.

The ISTQB CT-AI syllabus explains that data pre-processing includes activities such as data cleaning and data transformation. Data transformation covers changes to the format or representation of data, for example converting data types or encodings so that they can be correctly processed by machine learning algorithms.

Specifically, the syllabus states that transformation may include converting categorical values to numerical values or changing data representations, such as numbers stored as strings. This directly corresponds to Option D, which describes a dataset issue that can be resolved through pre-processing.

Option A is incorrect because insufficient data is a data availability problem that cannot be solved by pre-processing.
Option B is incorrect because invalid data is typically addressed through data cleaning rather than transformation.
Option C is incorrect because wanted outliers are intentionally retained and should not be removed or altered during pre-processing.

Thus, Option D is the correct answer.

The ISTQB Certified Tester AI Testing Syllabus v1.0 explicitly categorizes different AI implementation technologies in Section 1.4 - AI Technologies. Within this section, AI methods are grouped into categories, one of which is "Reasoning techniques." These reasoning techniques include rule engines, deductive classifiers, case-based reasoning, and procedural reasoning. Because deductive classifiers are explicitly listed under this category, they are recognized as reasoning-based AI technologies.

Reasoning techniques differ from machine learning approaches because they rely on structured, predefined rules or logic to reach conclusions. Deductive classifiers use logical inference and symbolic reasoning to classify inputs by applying encoded knowledge, rather than learning patterns statistically from data.

The other options---linear regression, Random Forest, and genetic algorithms---are categorized in the syllabus as machine learning or optimization techniques, not reasoning techniques. Linear regression is a statistical prediction method, Random Forest is an ensemble-based machine learning model, and genetic algorithms are optimization approaches inspired by evolutionary processes.

Thus, based on the syllabus definitions, Deductive classifiers (Option A) are the only technology considered to be a reasoning technique.

The ISTQB CT-AI syllabus defines clustering as an unsupervised learning approach in which similarities among input data points are identified in order to group them based on common characteristics or attributes. The syllabus explicitly notes that clustering is often used in marketing scenarios, such as grouping customers according to behavior or tendencies.

Option A directly matches this definition, as associating shoppers with their shopping tendencies involves identifying natural groupings in customer behavior without predefined labels.

Option B is incorrect because grouping fish by fin types implies predefined categories, which aligns more with classification than clustering.
Option C is also a classification problem, as it involves assigning purchases to predefined categories based on perceived attractiveness.
Option D describes a prediction task involving numerical estimation, which is a regression problem rather than clustering.

Therefore, Option A is the correct example of a clustering problem solvable using unsupervised learning.