MLS-C01 Machine Learning Implementation and Operations Practice Test 12 — 15 Questions

Question 1

A company uses Amazon SageMaker to train models. The data scientist wants to automate the retraining process whenever new data arrives in an S3 bucket. Which THREE services can be used together to achieve this? (Choose THREE.)

Accepted Answer

Amazon S3. Options A, C, and D are correct. A: Amazon S3 can trigger events on new data. C: AWS Lambda can process the event and start the training job. D: SageMaker can run the training job. Option B (Amazon EC2) is not needed. Option E (AWS Glue) is for ETL, not directly for triggering retraining.

Answer

Amazon EC2

Answer

AWS Glue

Question 2

A company operates a real-time fraud detection system using an Amazon SageMaker endpoint. The model is a gradient boosting model trained on historical transaction data. The endpoint is deployed on an ml.c5.2xlarge instance with auto-scaling enabled based on average latency. Recently, during a flash sale event, the endpoint started returning HTTP 503 errors. The CloudWatch metrics show that the CPU utilization is at 70%, and the average latency has increased from 50 ms to 200 ms. The auto-scaling policy is configured to add one instance when average latency exceeds 100 ms for 5 consecutive minutes, and remove one instance when latency drops below 50 ms for 5 minutes. The current number of instances is 2. The flash sale lasted 30 minutes. What should the company do to prevent this issue in future flash sales?

Accepted Answer

Pre-warm the endpoint by setting a minimum number of instances that can handle the expected peak load before the flash sale. Option C is correct because the auto-scaling policy is reactive and too slow (5-minute evaluation period) to handle rapid traffic spikes. Pre-warming the endpoint by increasing the number of instances before the flash sale ensures capacity is available. Option A (increase instance size) may help but is more expensive and still reactive. Option B (use a different algorithm) is not the core issue. Option D (enable throttling) would still result in errors.

Answer

Enable request throttling to drop excess requests

Answer

Change the instance type to ml.c5.4xlarge to handle higher load

Answer

Change the model to a simpler model with lower latency

Question 3

A data scientist is using Amazon SageMaker to train a model on a large dataset (10 TB) stored in S3 in Parquet format. The training job uses an ml.p3.16xlarge instance with multiple GPUs. The data scientist notices that the GPU utilization is low (around 30%) and the training is slow. The dataset consists of hundreds of thousands of small Parquet files. The data scientist suspects that the I/O is bottlenecked. What should the data scientist do to improve GPU utilization and training speed?

Accepted Answer

Consolidate the small Parquet files into larger files (e.g., 1 GB each). Option A is correct because consolidating small files into larger files reduces the overhead of reading many files from S3, improving I/O throughput and keeping GPUs busy. Option B (use Pipe mode) may help but does not address the file size issue. Option C (increase batch size) may improve utilization but the I/O bottleneck remains. Option D (use a smaller instance) would not improve speed.

Answer

Increase the batch size

Answer

Use a smaller instance type to reduce cost

Answer

Use Pipe input mode to stream data directly

Question 4

A company has deployed a machine learning model on a SageMaker endpoint that serves predictions to a web application. The model uses a custom inference container that loads the model artifacts from an ECR repository. After updating the model with new training data, the data scientist creates a new model and updates the endpoint. However, some users report that they still get predictions from the old model. The data scientist confirms that the endpoint configuration points to the new model. What is the most likely cause?

Accepted Answer

DNS caching on the client side is resolving to the old endpoint IP address. Option B is correct because DNS caching at the client side may resolve the endpoint's DNS name to an old IP address, especially if the endpoint's underlying instances have not changed. Option A (incorrect model artifact) would affect all users. Option C (load balancer) is not part of SageMaker. Option D (CloudFront caching) could be an issue if CloudFront is in front, but the question does not mention it.

Answer

The new model artifacts are not correctly uploaded to S3

Answer

The endpoint is behind a load balancer that is not updated

Answer

The inference container is cached and not pulling the new image

Question 5

A data scientist is using SageMaker to train a model. The training job is failing with a 'ResourceLimitExceeded' error. Which action should be taken to resolve this issue?

Accepted Answer

Request a service limit increase for the instance type.. Option B is correct because the error indicates that the account's maximum number of instances for the instance type has been reached. Requesting a limit increase for the specific instance type is the appropriate resolution. Option A is incorrect because it does not address the limit issue. Option C is incorrect because switching to a different instance type may not be desired and does not resolve the underlying limit. Option D is incorrect because the error is about resource limits, not data size.

Answer

Use a different AWS Region.

Answer

Reduce the training dataset size.

Answer

Switch to a different instance type with lower resource requirements.

Question 6

A company is deploying a real-time inference endpoint using SageMaker. The model has a high memory footprint and requires GPU acceleration. Which instance type and configuration should be used to minimize cost while meeting latency requirements?

Accepted Answer

ml.p3.2xlarge with 1 GPU. Option C is correct because the p3.2xlarge provides GPU acceleration with sufficient memory for a high-memory model and is cost-effective for real-time inference. Option A is incorrect because ml.c5.xlarge does not have GPU. Option B is incorrect because p3.16xlarge is too large and expensive. Option D is incorrect because ml.g4dn.xlarge has less memory and may not be suitable for high memory footprint.

Answer

ml.g4dn.xlarge with 1 GPU

Answer

ml.c5.xlarge with no GPU

Answer

ml.p3.16xlarge with 8 GPUs

Question 7

A machine learning team is using AWS Glue to prepare data for training. They notice that the ETL job takes a long time to process large datasets. Which change is most likely to improve performance?

Accepted Answer

Increase the number of DPUs for the Glue job.. Option A is correct because increasing the number of DPUs (Data Processing Units) in AWS Glue can parallelize processing and reduce job duration. Option B is incorrect as it may reduce parallelism. Option C is incorrect because Spark shuffle is necessary; avoiding it may not be feasible. Option D is incorrect because reducing data size is not always possible.

Answer

Decrease the number of workers in the Glue job.

Answer

Disable Spark shuffle operations.

Answer

Reduce the dataset size by sampling.

Question 8

A company is using SageMaker Ground Truth to label images for a computer vision model. After launching the labeling job, they notice that the labeling throughput is lower than expected. What should they do to increase throughput?

Accepted Answer

Use a private workforce with more workers.. Option D is correct because using a private workforce with more workers increases parallelism. Option A is incorrect because reducing the number of workers decreases throughput. Option B is incorrect because using a single annotator per task may not speed up labeling. Option C is incorrect because increasing task time per image slows down throughput.

Answer

Change the labeling task to use a single annotator per image.

Answer

Reduce the number of workers assigned to each task.

Answer

Increase the time allowed for each labeling task.

Question 9

A data scientist is using SageMaker Debugger to monitor a training job. The training loss is not decreasing as expected. Which Debugger feature can help identify the issue?

Accepted Answer

Built-in rules to detect training anomalies. Option B is correct because Debugger's built-in rules can detect vanishing gradients, overfitting, etc. Option A is incorrect because Debugger does not automatically tune hyperparameters. Option C is incorrect because saving tensors does not identify the issue automatically. Option D is incorrect because Debugger is not for model deployment.

Answer

Automatic hyperparameter tuning

Answer

Saving tensors every step

Answer

Deploying a model endpoint for real-time monitoring

Question 10

A company is using Amazon Rekognition to detect objects in images stored in S3. They want to reduce costs by processing images only when they are uploaded. Which AWS service should be used to trigger Rekognition automatically?

Accepted Answer

AWS Lambda. Option C is correct because S3 events can trigger Lambda, which calls Rekognition API. Option A is incorrect because CloudWatch Events can trigger on a schedule, not directly on S3 uploads. Option B is incorrect because Step Functions can orchestrate but not directly triggered by S3 uploads without Lambda. Option D is incorrect because SNS is passive and needs a subscriber.

Answer

Amazon CloudWatch Events

Answer

Amazon Simple Notification Service (SNS)

Answer

AWS Step Functions

Question 11

A machine learning engineer needs to deploy a model that requires custom inference code with dependencies. Which SageMaker deployment option should be used?

Accepted Answer

Create a custom Docker container and deploy to SageMaker endpoint.. Option A is correct because a custom container allows full control over dependencies and inference code. Option B is incorrect because a built-in algorithm may not support custom code. Option C is incorrect because a SageMaker notebook instance is for development, not deployment. Option D is incorrect because a batch transform job is for batch inference, not real-time.

Answer

Use a SageMaker notebook instance as an endpoint.

Answer

Use a built-in SageMaker algorithm.

Answer

Use a SageMaker batch transform job.

Question 12

A company is training a deep learning model on SageMaker using multiple GPUs. The training is slow due to inefficient data loading. Which TWO actions can improve I/O performance?

Accepted Answer

Use Pipe mode input for training data.. Options A and D are correct. Using Pipe mode streams data directly from S3, avoiding local disk writes. Using Amazon EFS can also provide shared, high-throughput storage. Option B is incorrect because SageMaker does not support EBS volumes as input directly; you would need to use FSx or EFS. Option C is incorrect because instance store volumes are ephemeral and not suitable for persistent data. Option E is incorrect because a single large instance may not improve I/O parallelism.

Answer

Use instance store volumes for data.

Answer

Increase the instance count to a single large instance.

Answer

Use Amazon EBS volumes attached to training instances.

Question 13

A data scientist is using SageMaker to build a model for fraud detection. The dataset is highly imbalanced. Which THREE techniques should be applied to address class imbalance?

Accepted Answer

Apply SMOTE to generate synthetic samples of the minority class.. Options A, C, and E are correct. Oversampling the minority class (e.g., SMOTE) and undersampling the majority class help balance the dataset. Using class weights in the loss function penalizes misclassifications of minority class more. Option B is incorrect because using accuracy as the metric can be misleading for imbalanced datasets; precision-recall or AUC is better. Option D is incorrect because training only on majority class would ignore minority class entirely.

Answer

Train the model only on the majority class.

Answer

Use accuracy as the evaluation metric.

Question 14

A company is using SageMaker Autopilot to automatically build ML models. They want to ensure that the generated models are reproducible. Which TWO settings should they configure?

Accepted Answer

Set a random seed.. Options B and D are correct. Setting a random seed ensures that random processes (like train/test split) are deterministic. Disabling automatic data splitting and manually providing a validation split gives control over the split. Option A is incorrect because enabling early stopping can vary based on training dynamics. Option C is incorrect because using multiple trials introduces randomness. Option E is incorrect because automatic feature engineering may introduce non-deterministic transforms.

Answer

Use multiple trials.

Answer

Enable early stopping.

Answer

Enable automatic feature engineering.

Question 15

A company operates a real-time fraud detection system using SageMaker. The model is deployed on an ml.c5.xlarge instance behind an Application Load Balancer (ALB). Recently, during a sales event, traffic spiked and the endpoint returned HTTP 503 errors. The team scaled the instance count from 2 to 5, but errors persisted. CloudWatch metrics show low CPU utilization (~30%) and high memory usage (~90%). The model loads a large dictionary file (2GB) into memory at startup. Which action should resolve the issue?

Accepted Answer

Use a memory-optimized instance type like r5.large.. Option C is correct because the high memory usage indicates the instance type does not have enough memory to handle concurrent requests. Using a memory-optimized instance like r5.large provides more memory per vCPU, reducing memory pressure and preventing OOM errors. Option A is incorrect because low CPU utilization suggests the bottleneck is not CPU. Option B is incorrect because increasing instances does not help if each instance is memory-constrained. Option D is incorrect because using Spot instances may cause interruptions and does not fix the memory issue.

Answer

Enable auto-scaling with Spot instances.

Answer

Switch to a compute-optimized instance type like c5.2xlarge.

Answer

Increase the number of instances further to 10.