AI0-001 Practice Test 2 — 10 Questions

Question 1

An AI model is trained to predict loan default. The training data contains 95% non-default and 5% default. Which metric is most appropriate to evaluate model performance given the imbalanced dataset?

Accepted Answer

F1-score. The F1-score is the harmonic mean of precision and recall, making it robust to class imbalance. In this dataset with 95% non-default and 5% default, accuracy would be misleadingly high (95%) even if the model never predicts default, while F1-score penalizes poor recall of the minority class.

Answer

Mean squared error

Answer

Accuracy

Answer

R-squared

Question 2

Which TWO of the following are best practices for securing an AI model against adversarial attacks?

Accepted Answer

Adversarial training with perturbed examples.. Adversarial training explicitly augments the training dataset with perturbed examples (e.g., using FGSM or PGD attacks) to teach the model to recognize and resist malicious inputs. This method directly hardens the model against evasion attacks by improving its decision boundary robustness.

Answer

Model pruning to reduce the number of parameters.

Answer

Increasing model complexity to capture more patterns.

Answer

Hyperparameter optimization using grid search.

Question 3

An organization is deploying a machine learning model that classifies loan applications. They want to prevent an attacker from reconstructing individual customer records from the model's predictions. Which type of attack should they defend against?

Accepted Answer

Model inversion. Model inversion attacks allow an attacker to reconstruct the original training data by analyzing the model's predictions. In this scenario, the attacker could use the model's outputs to infer sensitive details about individual loan applicants, such as income or credit history, violating privacy. Defending against model inversion is critical when predictions can be used to reverse-engineer private training records.

Answer

Membership inference

Answer

Data poisoning

Answer

Adversarial example

Question 4

A company needs to store large volumes of unstructured data (PDFs, images, logs) for future AI model training. The data must be easily accessible by data scientists using Spark and must support cost-effective storage. Which data infrastructure is MOST appropriate?

Accepted Answer

Amazon S3 data lake. A data lake stores raw, unstructured data at low cost and integrates with Spark. Data warehouses are for structured, processed data; vector databases are for embeddings.

Answer

Snowflake data warehouse

Answer

Relational database like Amazon RDS

Answer

Pinecone vector database

Question 5

A self-driving car company is testing an AI model for pedestrian detection. During simulation, the model fails to detect pedestrians in low-light conditions. The safety team wants to improve robustness without retraining the entire model from scratch. Which approach is most appropriate?

Accepted Answer

Apply data augmentation techniques to simulate low-light conditions in the training dataset.. Data augmentation techniques, such as adjusting brightness, contrast, and adding noise, can synthetically create low-light training examples from existing data. This improves the model's robustness to low-light conditions without requiring a full retraining from scratch, as it directly addresses the distribution shift in the input data.

Answer

Replace the convolutional layers with transformer layers to improve attention.

Answer

Use adversarial training to add imperceptible perturbations to training images.

Answer

Increase the model's depth by adding more convolutional layers.

Question 6

A data scientist is training a model to detect fraudulent transactions. To protect customer privacy, the team wants to ensure that the model does not inadvertently memorize and reveal sensitive information about individuals in the training set. Which technique should be applied during training?

Accepted Answer

Differential privacy. Differential privacy is the correct technique because it adds calibrated noise to the training process or output, ensuring that the model cannot infer whether any specific individual's data was included in the training set. This directly addresses the goal of preventing memorization and leakage of sensitive information while still allowing the model to learn useful patterns for fraud detection.

Answer

Federated learning

Answer

Homomorphic encryption

Answer

Model quantization

Question 7

An organization wants to implement an AI ethics board. Which composition best ensures independence and expertise?

Accepted Answer

Mix of internal stakeholders and external ethicists. An AI ethics board must combine internal stakeholders (who understand organizational context, data flows, and operational constraints) with external ethicists (who provide independent, unbiased perspectives and specialized knowledge of ethical frameworks like IEEE Ethically Aligned Design or the EU AI Act). This composition ensures the board can evaluate AI systems for bias, fairness, and transparency without being dominated by business or technical interests, which is critical for maintaining trust and regulatory compliance.

Answer

All members from the legal department

Answer

IT department head and data scientists

Answer

Only senior executives from the company

Question 8

A retail company uses a gradient boosting model to predict customer lifetime value (CLV). The model currently uses 50 features including purchase history, demographics, and web behavior. The model's RMSE on the test set is 120. The data science team wants to improve the model's accuracy without increasing training time significantly. They have access to additional data: customer support interaction logs (text), social media sentiment (text), and third-party credit scores (numeric). They also have the ability to perform feature engineering, hyperparameter tuning, and ensemble methods. Which approach is most likely to yield the best improvement in predictive performance with minimal increase in training time?

Accepted Answer

Engineer new features such as average purchase value and recency. Engineering domain-relevant features like average purchase value and recency directly captures the underlying behavioral patterns that drive customer lifetime value, often providing a higher signal-to-noise ratio than adding raw text or third-party data. This approach leverages existing data without significantly increasing the feature dimensionality or training time, unlike adding TF-IDF vectors which would dramatically expand the feature space and slow training.

Answer

Add the customer support text as a feature using TF-IDF vectors

Answer

Use an ensemble of gradient boosting and random forest models

Answer

Perform hyperparameter tuning using grid search

Question 9

During model training, the data science team discovers that many input features contain missing values. Which step should be taken to improve data quality?

Accepted Answer

Implement data validation checks to handle missing data appropriately (e.g., imputation).. Data validation checks, such as imputation (e.g., mean, median, or KNN imputation), directly address missing values by estimating plausible replacements based on the available data. This improves data quality and prevents bias or loss of information that could degrade model performance. In the context of AI implementation, handling missing data is a fundamental data preprocessing step to ensure robust model training.

Answer

Increase the model complexity to handle missing data.

Answer

Ignore missing values and train the model.

Answer

Remove all records with missing values.

Question 10

A team is using an API from a cloud AI service to generate text. They notice that repeated requests with the same prompt return different outputs. They want consistent responses for testing. Which parameter should they adjust?

Accepted Answer

Set the temperature to 0. Setting the temperature to 0 makes the model deterministic, producing the same output for the same input, which is ideal for testing.

Answer

Increase the top_p parameter to 1.0

Answer

Set the frequency_penalty to 0

Answer

Increase the max_tokens parameter