AWS Certified AI Practitioner (AIF-C01) — Practice Exam

Question 1

Nordhaven Freight runs a fleet of refrigerated trucks. Engineers log 14 sensor readings per truck per minute and have three years of historical records showing which compressor units eventually failed and on what date. Leadership wants a system that flags a compressor as likely to fail within the next 30 days. Which machine learning approach fits this problem?

Accepted Answer

Supervised learning, because the historical records provide labeled failure outcomes to train against

Answer

Unsupervised learning, because the sensor readings arrive as a continuous unlabeled stream

Answer

Semi-supervised learning, because sensor data is numeric rather than categorical

Answer

Reinforcement learning, because the trucks operate in a changing physical environment

Question 2

A data scientist at Kestrel Analytics trains a customer-churn model that scores 98.4% accuracy on the training set but only 71% on a held-out test set. Which description best characterizes what has happened and the most direct remedy?

Accepted Answer

The model is overfitting; the team should apply regularization or gather more diverse training data

Answer

The training data is imbalanced; the team should switch the loss function to mean squared error

Answer

The model is underfitting; the team should add more layers and train for more epochs

Answer

The features are poorly scaled; the team should convert all numeric columns to text embeddings

Question 3

Tilbury Municipal Water wants to group its 60,000 residential accounts into segments based on consumption patterns so that conservation messaging can be tailored. Nobody has defined what the segments should be, and no account carries a segment label. Which technique should the analytics team apply?

Accepted Answer

Clustering, such as k-means, over the consumption feature set

Answer

Linear regression to predict next month's consumption volume

Answer

Binary classification using logistic regression

Answer

Time-series forecasting with a seasonal decomposition model

Question 4

The claims department at Palladine Insurance built a model that decides whether a submitted photo shows hail damage. The team must report a single metric to the regulator that balances how many genuine hail claims the model catches against how many of its positive calls are correct. Which metric should the team report?

Accepted Answer

F1 score

Answer

Mean absolute percentage error

Answer

Root mean squared error

Answer

R-squared

Question 5

A retail chain called Brightwick Home wants to convert scanned supplier invoices into structured fields such as vendor name, invoice number, and line-item totals, without building or training any model. Which AWS service is designed for this task?

Accepted Answer

Amazon Textract

Answer

Amazon Polly

Answer

Amazon Kendra

Answer

Amazon Comprehend