MLA-C01 Question Bank · Latest 50 Questions Free Preview (Page 2/5)
Questions 11–20 of the latest 50 MLA-C01 questions (5 pages total), with answers and explanations, no login required. Practice all 233 questions online.
Question 11 · No. 194 · Single choice
A company is preparing data to train a new ML model on Amazon SageMaker AI. The data has not been used before for ML training. The data includes duplicates and is missing some values. The company needs to increase the data quality and detect any statistical bias in the data. Which solution will meet these requirements?
- A. Use SageMaker Clarify to create data quality rules. Use SageMaker Model Monitor to detect bias.
- B. Use SageMaker Data Wrangler to create data quality rules. Use SageMaker Clarify to detect bias.
- C. Use SageMaker Debugger to create data quality rules. Use SageMaker Model Monitor to detect bias.
- D. Use SageMaker Model Monitor to create data quality rules. Use SageMaker Clarify to detect bias.
Answer: B
Explanation: SageMaker Data Wrangler is specifically designed for data preparation tasks and can create data quality rules to handle issues like duplicates and missing values. SageMaker Clarify is the appropriate tool for detecting statistical bias in datasets before training. SageMaker Model Monitor is used for monitoring deployed models in production, and SageMaker Debugger is used for debugging training jobs, making options A, C, and D incorrect.
Question 12 · No. 195 · Single choice
An ML engineer uses A/B testing to dynamically select recommendation models. The models are deployed on Amazon SageMaker AI endpoints. The ML engineer needs to monitor system metrics such as latency, call volume, and HTTP status codes when the endpoints are invoked. Which solution will meet these requirements with the LEAST operational overhead?
- A. Use AWS X-Ray tracing to monitor SageMaker AI endpoints.
- B. Configure Amazon CloudWatch dashboards with AWS Lambda log processors.
- C. Enable AWS Identity and Access Management (IAM) Access Analyzer to track SageMaker AI endpoint metrics.
- D. Deploy AWS Trusted Advisor checks on SageMaker AI endpoint instances.
Answer: B
Explanation: Amazon CloudWatch is the native AWS monitoring service that integrates seamlessly with SageMaker AI endpoints. It automatically collects metrics such as latency, invocation count (call volume), and can monitor HTTP status codes through CloudWatch Logs and metrics. Even though option B mentions Lambda log processors, CloudWatch remains the most operationally efficient solution among the choices. Option A (X-Ray) is primarily for distributed tracing of individual requests, not aggregate system metrics. Option C (IAM Access Analyzer) is for analyzing IAM permissions, not monitoring metrics. Option D (Trusted Advisor) provides best-practice recommendations and cost optimization, not real-time endpoint metrics. Therefore, B is the best answer for monitoring the specified metrics with the least operational overhead.
Question 13 · No. 196 · Single choice
A company is using Amazon SageMaker AI to build an ML model to predict customer behavior. The company needs to explain the bias in the model to an auditor. The explanation must focus on demographics data of the customers. Which solution will meet these requirements?
- A. Use SageMaker Clarify to generate a bias report. Send the report to the auditor.
- B. Use AWS Glue DataBrew to create a job to detect drift in the model's data quality. Send the job output to the auditor.
- C. Use Amazon Quick Suite (previously known as Amazon QuickSight) integration with SageMaker AI to generate a bias report from Quick Suite. Send the report to the auditor.
- D. Use Amazon CloudWatch metrics from the SageMaker AI namespace to create a bias dashboard. Share the dashboard with the auditor.
Answer: A
Explanation: Amazon SageMaker Clarify is specifically designed to detect bias in machine learning models, with a focus on demographic data such as age, gender, and race. It provides bias metrics and generates detailed bias reports that can be shared with auditors. The other options are not suitable: Glue DataBrew focuses on data quality and drift detection but not ML model bias on demographics; QuickSight is a BI visualization tool, not designed for ML bias detection; CloudWatch is for monitoring infrastructure metrics, not for analyzing ML model bias.
Question 14 · No. 197 · Single choice
An ML engineering team is spread across multiple locations. When the lead ML engineer opens an Amazon SageMaker Al notebook, the ML engineer does not see the latest merged notebook made by other team members from a Git repository. The lead ML engineer must see the latest SageMaker AI notebook updates. Which solution will meet this requirement?
- A. Run the !git pull origin master command.
- B. Run the !git commit command.
- C. Run the !git push origin master command.
- D. Run the !git branch command.
Answer: A
Explanation: The !git pull origin master command fetches the latest changes from the remote master branch and merges them into the local repository. This will update the SageMaker AI notebook environment with the latest merged notebook updates made by other team members. The other options are incorrect: git commit only saves local changes, git push sends local changes to remote (opposite direction), and git branch only lists available branches without updating the local repository.
Question 15 · No. 198 · Single choice
A recommendation model uses ML and calls an Amazon SageMaker AI endpoint to get recommendations. An ML engineer must ensure that the model stays available during an expected increase in user traffic. Which solution will meet these requirements?
- A. Configure auto scaling on the SageMaker AI endpoint.
- B. Create a new SageMaker AI endpoint. Deploy the model to the new endpoint.
- C. Use SageMaker Neo to optimize the model for inference.
- D. Attach an Auto Scaling group to the SageMaker AI endpoint.
Answer: A
Explanation: The correct solution is to configure auto scaling on the SageMaker AI endpoint. Amazon SageMaker provides built-in auto scaling capabilities that automatically adjust the number of instances serving the endpoint based on traffic patterns and defined policies (such as target tracking or step scaling). This ensures the model remains available during expected increases in user traffic. Option B (creating a new endpoint) does not address scaling and would not improve the situation. Option C (SageMaker Neo) optimizes model performance for inference but does not handle traffic spikes or availability. Option D is incorrect because Auto Scaling groups are EC2 features and cannot be directly attached to SageMaker endpoints—SageMaker has its own native auto scaling functionality.
Question 16 · No. 199 · Single choice
A company wants to use large language models (LLMs) that are supported by Amazon Bedrock to develop a chat interface for the company’s internal technical documentation. The company stores the documentation as dozens of text files that are several megabytes in total size. The company updates the text files often. Which solution will meet these requirements MOST cost-effectively?
- A. Create a new LLM on Amazon Bedrock. Train the new LLM on the original dataset and the company documentation. Make the new model available in Bedrock for calls from the chat interface.
- B. Integrate the company documentation with Amazon Bedrock guardrails. Invoke the guardrails for all Amazon Bedrock calls from the chat interface.
- C. Use all the text files to fine tune a model in Amazon Bedrock. Use the fine-tuned model to process user prompts.
- D. Upload all the text files to an Amazon Bedrock knowledge base. Use the knowledge base to provide context when the chat interface makes calls to Amazon Bedrock.
Answer: D
Explanation: An Amazon Bedrock knowledge base uses retrieval-augmented generation (RAG) to index the documentation and retrieve relevant context for each prompt. The company can update or replace documents without retraining or fine-tuning a model, making it the most cost-effective solution. Guardrails manage model behavior but do not provide document retrieval, while creating or fine-tuning a model would be unnecessary and more expensive for frequently changing content.
Question 17 · No. 200 · Single choice
An ML engineer at an insurance company trains a regression model to predict the number of insurance policy sales each month. After training the model, the ML engineer uses Amazon SageMaker AI to deploy the model for inference. The ML engineer wants to monitor the model predictions to detect whether the production data distribution differs from the training data distribution when there are changes in customer behaviors. Which solution will meet these requirements?
- A. Determine whether there is drift in the data quality.
- B. Determine whether there is drift in the model quality.
- C. Determine whether there is drift in the model bias.
- D. Determine whether there is drift in the feature attribution.
Answer: A
Explanation: The requirement is to detect whether the production data distribution differs from the training data distribution due to changes in customer behaviors. This is a classic case of data drift detection. Amazon SageMaker's Data Quality monitoring (Model Monitor) tracks statistical properties of input features and compares them between training and production datasets, detecting when customer behavior changes cause distribution shifts. Model quality (B) requires ground truth labels and monitors prediction performance metrics, not data distribution. Model bias (C) tracks fairness metrics across sensitive attributes. Feature attribution (D) monitors changes in feature importance via SHAP values, not input distribution itself.
Question 18 · No. 201 · Single choice
A company is performing A/B testing on a model that provides product recommendations. The company has deployed two versions of the model and is showing each version of the model to 50% of users randomly. Which metric should the company use to evaluate whether users act on a recommendation?
- A. The conversion rates between the two versions of the model
- B. The number of recommendations provided to each user
- C. The model accuracy on held-out test data
- D. The latency of model inference
Answer: A
Explanation: Conversion rates directly measure whether users take action on the recommendations (e.g., clicking, purchasing, or engaging with the suggested product). In A/B testing for recommendation models, conversion rate is the key business metric that captures user behavior in response to the recommendations. The other options either measure irrelevant aspects (number of recommendations, latency) or are pre-deployment metrics (accuracy on held-out test data) that don't reflect real-world user actions.
Question 19 · No. 202 · Multiple choice
A company is developing an application that reads animal descriptions from user prompts and generates images based on the information from the prompts. The application reads a message from an Amazon Simple Queue Service (Amazon SQS) queue. Then the application uses Amazon Titan Image Generator on Amazon Bedrock to generate an image based on the information in the message. Finally, the application removes the message from SQS queue. Which IAM permissions should the company assign to the application’s IAM role? (Choose two.)
- A. Allow the bedrock:InvokeModel action for the Amazon Titan Image Generator resource.
- B. Allow the bedrock:Get* action for the Amazon Titan Image Generator resource.
- C. Allow the sqs:ReceiveMessage action and the sqs:DeleteMessage action for the SQS queue resource.
- D. Allow the sqs:GetQueueAttributes action and the sqs:DeleteMessage action for the SQS queue resource.
- E. Allow the sagemaker:PutRecord* action for the Amazon Titan Image Generator resource.
Answer: A, C
Explanation: Option A is correct because to invoke Amazon Titan Image Generator on Amazon Bedrock, the application needs the bedrock:InvokeModel permission. Option C is correct because the application needs sqs:ReceiveMessage to read messages from the SQS queue and sqs:DeleteMessage to remove the message from the queue after processing. Option B is incorrect because Get* actions are for retrieving information, not for invoking models. Option D is incorrect because GetQueueAttributes is not used to read messages; ReceiveMessage is required. Option E is incorrect because Amazon Titan Image Generator is on Amazon Bedrock, not SageMaker, and PutRecord is for Kinesis streams.
Question 20 · No. 203 · Single choice
A company is creating an ML model to identify defects in a product. The company has gathered a dataset and has stored the dataset in TIFF format in Amazon S3. The dataset contains 200 images in which the most common defects are visible. The dataset also contains 1,800 images in which there is no defect visible. An ML engineer trains the model and notices poor performance in some classes. The ML engineer identifies a class imbalance problem in the dataset. What should the ML engineer do to solves this problem?
- A. Use a few hundred images and Amazon Rekognition Custom Labels to train a new model.
- B. Undersample the 200 images in which the most common defects are visible.
- C. Oversample the 200 images in which the most common defects are visible.
- D. Use all 2,000 images and Amazon Rekognition Custom Labels to train a new model.
Answer: C
Explanation: The dataset has a significant class imbalance with only 200 images containing defects versus 1,800 images without defects (a 1:9 ratio). To address this, the ML engineer should oversample the minority class (the 200 defect images) by creating duplicates or augmentations of these images. This will help balance the dataset and improve the model's ability to learn the features of the defect class. Option B would make the imbalance worse, Options A and D do not address the class imbalance issue.