An AI model's documentation shows that the modeler utilized K-means clustering to categorize a data set that included a large volume of records. Which of the following is an advantage of using a K-means clustering algorithm for determining classification categories to group applicable data before it is inserted into the learning model?
Answer(s): A
The advantage of using K-means clustering is that it is an unsupervised algorithm, allowing data to be grouped without requiring labeled examples. This reduces subjectivity and potential bias from manual data labeling,making it effective for handling large volumes of records.
An anti-fraud team in a financial organization is building a machine learning (ML) model to detect online fraud, and an internal audit team has been engaged to audit the model. When auditing the model training stage, which of the following is MOST important to include in the scope of the audit?
Answer(s): B
Data fairness is the most important factor to audit in the training stage because biased or unbalanced training data can lead to discriminatory or inaccurate fraud detection. Ensuring fairness directly impacts the model’s reliability, ethical use, and compliance.
An IS auditor learns that the training data for an AI solution has insufficient representation. Which of the following would be the GREATEST risk in this situation?
The greatest risk with insufficient representation in training data is imbalance. When certain groups or categories are underrepresented, the AI model becomes biased toward the majority class, leading to unfair, inaccurate, or unreliable outputs.
Which of the following BEST improves predictions and understanding to refine the accuracy of AI models?
Data augmentation best improves predictions and understanding by increasing the diversity and volume of training data. This helps AI models generalize better, reduces overfitting, and refines overall prediction accuracy.
Which key performance indicator (KPI) BEST indicates the quality of the data used in AI model training?
Answer(s): D
The percentage of missing values removed from the data set is the best KPI for indicating data quality. High levels of missing data can compromise model training and accuracy, so tracking and minimizing missing values directly reflects data reliability.
Post your Comments and Discuss ISACA AAIA exam prep with other Community members:
Our website is free, but we have to fight against AI bots and content theft. We're sorry for the inconvenience caused by these security measures. You can access the rest of the AAIA content, but please register or login to continue.