QUESTION 61
An ML engineer is building a model to predict house and apartment prices. The model uses three features: Square Meters, Price, and Age of Building. The dataset has 10,000 data rows. The data includes data points for one large mansion and one extremely small apartment. The ML engineer must perform preprocessing on the dataset to ensure that the model produces accurate produces accurate predictions for the typical house or apartment. Which solution will meet these requirements?

A. Remove the outliers and perform a log transformation on the Square Meters variable.
B. Keep the outliers and perform normalization on the Square Meters variable.
C. Remove the outliers and perform one-hot encoding on the Square Meters variable.
D. Keep the outliers and perform one-hot encoding on the Square Meters variable.

Correct Answer: A

QUESTION 62
A company is developing an internal cost-estimation tool that uses an ML model in Amazon SageMaker AI. Users upload high-resolution images to the tool. The model must process each image and predict the cost of the object in the image. The model also must notify the user when processing is complete. Which solution will meet these requirements?

A. Store the images in an Amazon S3 bucket. Deploy the model on SageMaker AI. Use batch transform jobs for model inference. Use an Amazon Simple Queue Service (Amazon SQS) queue to notify users.
B. Store the images in an Amazon S3 bucket. Deploy the model on SageMaker AI. Use an asynchronous inference strategy for model inference. Use an Amazon Simple Notification Service (Amazon SNS) topic to notify users.
C. Store the images in an Amazon Elastic File System (Amazon EFS) file system. Deploy the model on SageMaker AI. Use batch transform jobs for model inference. Use an Amazon Simple Queue Service (Amazon SQS) queue to notify users.
D. Store the images in an Amazon Elastic File System (Amazon EFS) file system. Deploy the model on SageMaker Al. Use an asynchronous inference strategy for model inference. Use an Amazon Simple Notification Service (Amazon SNS) topic to notify users.

Correct Answer: B

QUESTION 63
A credit card company rained a fraud detection model on a dataset in which 3% of transactions are fraudulent. The company performed A/B testing after a recent model update Customers reported that an increased number of legitimate transactions were incorrectly fagged as fraudulent. The company trains and tests a new model The company must determine whether the new model will reduce the rate at which legitimate transactions are incorrectly flagged as fraudulent. The new model must maintain the same level of fraud detection accuracy as the original model. Which metric or metrics will indicate that the company should replace the original model with the new model?

A. The new model has better Area Under the Precision-Recall Curve (AUC-PR) than the original model
B. The new model has better recall that the original model and the same level of precision.
C. The new model has better precision than the original model and the same level of recall.
D. The new mode has better overall accuracy than the original model.

Correct Answer: C

QUESTION 64
An ML engineer is building an ML model in Amazon SageMaker AI. The ML engineer needs to load historical data directly from Amazon S3, Amazon Athena, and Snowflake into SageMaker AI. Which solution will meet this requirement?

A. Use AWS Glue DataBrew to import the data into SageMaker AI.
B. Build a pipeline in SageMaker Pipelines to process the data. Use AWS DataSync to load the processed data into SageMaker A.
C. Create a feature store in SageMaker Feature Store. Use an Apache Spark connector to Feature Store to access the data
D. Use SageMaker Data Wrangler to query and import the data.

Correct Answer: D

QUESTION 65
An ML engineer is using AWS CodeDeploy to deploy new versions of containers that perform inference to an Amazon Elastic Container Service (Amazon ECS) compute platform. The ML engineer needs to define a deployment configuration to shift traffic to the updated ECS task set. The deployment configuration must shift 10% of traffic to the updated containers in the first increment. The remaining 90% of traffic must shift to the updated containers within 10 to 15 minutes. Which deployment configuration will meet these requirements?

A. CodepeployDefault.LambdaLinear10PercentEvery10Minutes
B. CodeDeployDefault.ECSAIIAtOnce
C. CodeDeployDefault.ECSCanary10Percent15Minutes
D. CodeDeployDefault.LambdaCanary10Percent15Minutes

Correct Answer: C

QUESTION 66
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.

Correct Answer: B

QUESTION 67
An ML engineer is using AWS Glue to transform proprietary data from a third-party vendor to a format that the ML engineer intends to use with the Amazon SageMaker DeepAR forecasting algorithm. The data includes several similar time series data files that the ML engineer must convert to the appropriate format The ML engineer must compress the files to optimize storage costs. Which solution will meet these requirements?

A. Use Snappy to convert the files to RecordIO-Protobuf and to compress the files.
B. Use XZ to convert the files to RecordIO-Protobuf and to compress the files.
C. Use XZ to convert the files to Apache Parquet format and to compress the files.
D. Use gzip to convert the files to Apache Parquet and to compress the files.

Correct Answer: A

QUESTION 68
An ML engineer is deploying a generative Al model-based customer support agent that uses Amazon SageMaker Al for inference. The customer support agent must respond to customer questions about topics such as shipping policies, refund processes, and account management. The generative Al model generates one token at a time. Customers report dissatisfaction with how long the customer support agent takes to generate lengthy responses to questions. The ML engineer must apply an inference optimization technique to improve the performance of the customer support agent. Which solution will meet this requirement?

A. Compilation
B. Speculative decoding
C. Quantization
D. Fast model loading

Correct Answer: B

QUESTION 69
An ML engineer uses one ML framework to train multiple ML models. The ML engineer needs to optimize the inference costs and host the models on Amazon SageMaker AI. Which solution will meet these requirements MOST cost-effectively?

A. Create a multi-container inference endpoint for direct invocation.
B. Create a multi-model inference endpoint for all the models.
C. Create a multi-container inference endpoint for sequential invocation.
D. Create multiple single-model inference endpoints for each model.

Correct Answer: B

QUESTION 70
A company is building an enterprise AI platform. The company must catalog models for production, manage model versions, and associate metadata such as training metrics with models. The company needs to eliminate the burden of managing different versions of models. Which solution will meet these requirements?

A. Use the Amazon SageMaker Model Registry to catalog the models. Create unique tags for each model version. Create key-value pairs to maintain associated metadata.
B. Use the Amazon SageMaker Model Registry to catalog the models. Create model groups for each model to manage the model versions and to maintain associated metadata.
C. Create a separate Amazon Elastic Container Registry (Amazon ECR) repository for each model. Use the repositories to catalog the models and to manage model versions and associated metadata.
D. Create a separate Amazon Elastic Container Registry (Amazon ECR) repository for each model. Create unique tags for each model version. Create key-value pairs to maintain associated metadata.

Correct Answer: B

QUESTION 71
An ML engineer is building an ML pipeline. The pipeline must process a dataset in two ways by using Amazon Athena. The pipeline must use batch processing to perform large-scale data transformations and for model training. The pipeline must also use near real-time processing to perform low-latency queries for inference and analytics. Which file format will provide the LEAST latency for both types of processing?

A. CSV
B. Apache Parquet
C. Nested JSON
D. Deserialized JSON

Correct Answer: B

QUESTION 72
An ML engineer at a credit card company used Amazon SageMaker Al to build and deploy a new ML model. The model was trained on a copy of transactions from the production environment The copy of transactions contained only a few fraudulent transactions. After deployment to production, the model is underperforming. What should the ML engineer do to improve the model’s performance?

A. Retrain the model with a different built-in algorithm that is available in SageMaker Al.
B. Use random undersampling in SageMaker Data Wrangler to randomly reduce the majority class samples. Retrain the model.
C. Use Synthetic Minority Oversamping Technique (SMOTE) in SageMaker Data Wrangler to generate synthetic minority samples. Retrain the model.
D. Use random oversampling in SageMaker Data Wrangler to randomly duplicate minority samples. Retrain the model.

Correct Answer: C

QUESTION 73
A customer call center uses Amazon Transcribe to convert hundreds of audio recordings of conversations between customers and support agents to text files. The call center wants to use the text files to train an ML model. To comply with industry regulations, the call center must remove customer names, addresses, and phone numbers from the training text files. Which solution will meet these requirements with the LEAST development effort?

A. Use Amazon Bedrock Guardrails to process and redact personal information from the text files.
B. Use the AWS Glue Detect PII transform to remove personal information from the text files.
C. Store the text files in Amazon S3 buckets. Use S3 Object Lambda functions to redact personal information.
D. Configure an Amazon SageMaker Data Wrangler custom transformation to remove personal information from the text files.

Correct Answer: B

QUESTION 74
An ML engineer is training a text generation model on Amazon SageMaker AI. After several epochs, the loss function does not converge, and the model’s accuracy on the validation dataset starts to show oscillating results. The ML engineer needs to ensure that the model achieves generalization. Which solution will meet this requirement?

A. Increase the learning rate and decrease the mini-batch size.
B. Increase the learning rate as the number of epochs increases.
C. Decrease the learning rate and increase the mini-batch size.
D. Decrease the learning rate and decrease the mini-batch size.

Correct Answer: C

QUESTION 75
An ML engineer is using Amazon SageMaker Al to train a deep learning model. During testing, the model shows high variance. The ML engineer needs to adjust a hyperparameter to improve generalization without significantly increasing training time. Which solution will MOST improve the model generalization?

A. Increase the number of layers in the neural network.
B. Reduce the learning rate of the optimizer.
C. Increase the L2 regularization parameter.
D. Increase the batch size significantly.

Correct Answer: C

QUESTION 76
A company has a team of data scientists who use Amazon SageMaker AI notebook instances to test ML models. When the data scientists need new permissions, the company attaches the permissions to each individual role that was created during the creation of the SageMaker AI notebook instance. The company needs to centralize management of the team’s permissions. Which solution will meet this requirement?

A. Create a single IAM role that has the necessary permissions. Attach the role to each notebook instance that the team uses.
B. Create a single IAM group. Add the data scientists to the group. Associate the group with each notebook instance that the team uses.
C. Create a single IAM user. Attach the AdministratorAccess AWS managed IAM policy to the user. Configure each notebook instance to use the IAM user.
D. Create a single IAM group. Add the data scientists to the group. Create an IAM role. Attach the AdministratorAccess AWS managed IAM policy to the role. Associate the role with the group. Associate the group with each notebook instance that the team uses.

Correct Answer: A

QUESTION 77
An ML engineer is using an Amazon SageMaker Studio notebook to train a neural network by creating an estimator. The estimator runs a Python training script that uses Distributed Data Parallel (DDP) on a single instance that has more than one GPU. The ML engineer discovers that the training script is underutilizing GPU resources. The ML engineer must identify the point in the training script where resource utilization can be optimized. Which solution will meet this requirement?

A. Use Amazon CloudWatch metrics to create a report that describes GPU utilization over time.
B. Add SageMaker Profiler annotations to the training script. Run the script and generate a report from the results.
C. Use AWS CloudTrail to create a report that describes GPU utilization and GPU memory utilization over time.
D. Create a default monitor in Amazon SageMaker Model Monitor and suggest a baseline. Generate a report based on the constraints and statistics the monitor generates.

Correct Answer: B

QUESTION 78
An ML engineer needs to run intensive model training jobs each month that can take 48 to 72 hours to run. The training jobs can be interrupted and resumed without major issues. The ML engineer has a fixed budget and needs to optimize computing resources. Which solution will meet these requirements MOST cost-effectively?

A. Purchase Reserved Instances with a partial upfront payment.
B. Purchase On-Demand Instances with no commitment.
C. Purchase Amazon SageMaker Al Savings Plans.
D. Purchase Spot Instances that use automated checkpoints.

Correct Answer: D

QUESTION 79
An ML engineer observes that an ML model is overfitting the training data. Which actions should the ML engineer take to correct this problem? (Select TWO.)

A. Apply regularization on the model.
B. Apply Synthetic Minority Oversampling Technique (SMOTE).
C. Increase the number of features.
D. Decrease the number of features.
E. Train the model with more epochs.

Correct Answer: AD

QUESTION 80
A hospital is developing an internal tool that will use a large language model (LLM). Once a week, the tool must read several hundred medical cases from the previous week and create a summary of the cases. Which solution will meet these requirements MOST cost-effectively?

A. Host an LLM on Amazon Bedrock. Configure on-demand throughput.
B. Host an LLM on Amazon Bedrock. Configure provisioned throughput.
C. Use Amazon SageMaker JumpStart to host an LLM that runs a SageMaker AI transform job to process all medical records simultaneously.
D. Deploy an LLM from Amazon SageMaker JumpStart to a SageMaker AI Serverless Inference endpoint.

Correct Answer: A