MLS-C01 Practice Test 37 — 15 Questions

Question 1

A data engineer needs to transform and move 2 TB of data from an Amazon RDS for PostgreSQL instance to Amazon S3 daily. The transformation includes filtering, joining with data in S3, and aggregating. Which AWS services can be used together to accomplish this with minimal operational overhead? (Choose THREE.)

Choose an answer to begin — your selection is scored in the full session.

Accepted Answer

Amazon S3. Amazon S3 is correct because it serves as the target storage location for the transformed data. The daily 2 TB output from the ETL pipeline must be stored durably and cost-effectively, and S3 provides the ideal object storage layer for this purpose, especially when combined with AWS Glue for the transformation logic.

Answer

Amazon EMR

Answer

Amazon Redshift

Question 2

A data scientist is analyzing a dataset with numerical features and a binary target variable. The data scientist creates a pairplot and notices that one feature has a bimodal distribution when colored by the target class. What does this observation suggest?

Accepted Answer

The feature is likely predictive of the target.. A bimodal distribution separated by class indicates the feature can help distinguish between the classes, making it predictive. Option A is wrong because bimodality separated by class suggests the feature is useful, not irrelevant. Option C is wrong because bimodality is not an indication of outliers; it shows a pattern related to the target. Option D is wrong because bimodality does not imply missing values.

Answer

The feature is irrelevant and should be removed.

Answer

The feature contains outliers that need to be removed.

Answer

The feature has missing values that need to be imputed.

Question 3

A company uses Amazon SageMaker to train a model for fraud detection. The training data is highly imbalanced. The data scientist uses SMOTE to oversample the minority class. However, the model still has poor recall on the minority class. Which additional technique should the data scientist consider?

Accepted Answer

Use class weights in the loss function. SMOTE generates synthetic samples for the minority class, but it does not directly address the model's tendency to prioritize the majority class during training. By assigning higher class weights to the minority class in the loss function, the model penalizes misclassifications of minority samples more heavily, which directly improves recall on that class. This technique is especially effective when combined with oversampling, as it forces the optimizer to focus on the underrepresented class during gradient updates.

Answer

One-vs-rest encoding

Answer

L1 regularization

Answer

Principal component analysis (PCA)

Question 4

A data scientist wants to understand the relationship between a categorical feature with 3 levels and a continuous target variable. Which visualization is most appropriate?

Accepted Answer

Box plot grouped by category. A box plot grouped by category (Option C) is the most appropriate visualization because it directly compares the distribution of a continuous target variable across the three levels of a categorical feature. It displays median, quartiles, and potential outliers for each group, making it ideal for understanding central tendency, spread, and skewness in a side-by-side comparison.

Answer

Correlation matrix

Answer

Line chart

Answer

Scatter plot

Question 5

Which THREE techniques can help reduce overfitting in a neural network? (Choose 3)

Accepted Answer

Dropout. Dropout is a regularization technique that randomly drops a fraction of neurons during training, which prevents the network from relying too heavily on any single neuron and forces it to learn more robust features. This reduces overfitting by introducing noise that improves generalization.

Answer

Increasing the number of layers

Answer

Using a larger learning rate

Question 6

A company is using Amazon SageMaker to host a model that performs real-time inference. The model receives around 100 requests per second with occasional spikes up to 500 requests per second. The current endpoint uses 2 ml.m5.large instances. During spikes, latency increases significantly, and some requests time out. What is the MOST cost-effective solution to handle the spikes without losing requests?

Accepted Answer

Configure SageMaker managed scaling with a target tracking policy and add a buffer based on the average spike duration. SageMaker managed scaling with a target tracking policy automatically adjusts the number of instances based on a specified metric (e.g., invocation count or latency), and adding a buffer based on the average spike duration ensures that additional capacity is provisioned before the spike hits, preventing timeouts. This is the most cost-effective approach as it scales out during spikes and scales in during normal load, avoiding over-provisioning.

Answer

Replace the instances with a single larger instance type, such as ml.m5.4xlarge

Answer

Use an Amazon SQS queue to buffer incoming requests and process them asynchronously

Answer

Use AWS Lambda with a provisioned concurrency to handle the spikes

Question 7

Refer to the exhibit. A data scientist is unable to query a table in Amazon Athena that is located in the 'my-data-bucket' S3 bucket. The IAM policy shown is attached to the scientist's role. What is the most likely reason for the failure?

Accepted Answer

The policy does not allow s3:PutObject to write query results to an S3 bucket.. Athena queries require permissions to write query results to an S3 bucket, typically via 's3:PutObject' on an output location. The policy only allows 's3:GetObject' and 's3:ListBucket' on the data bucket, but lacks any 's3:PutObject' permission, causing the failure. Option A is incorrect because the policy does not reference any KMS actions. Option B is incorrect because 'athena:StartQueryExecution' is implicitly allowed (not denied), though it doesn't appear in the policy; Athena actions are not shown but the failure is due to S3 write permissions. Option C is incorrect because 's3:GetObject' is explicitly allowed on the bucket.

Answer

The policy does not allow decrypting data encrypted with AWS KMS.

Answer

The policy does not allow athena:StartQueryExecution.

Answer

The policy does not allow s3:GetObject on the bucket.

Question 8

A company is deploying a machine learning model for real-time fraud detection. The model must have low latency (under 100 ms) and high throughput. The data scientist trains a gradient boosting model and deploys it to a SageMaker endpoint with a single ml.c5.xlarge instance. During load testing, the endpoint exceeds the latency threshold. Which change is MOST likely to reduce latency?

Accepted Answer

Replace the model with a simpler model, such as logistic regression. Replacing the gradient boosting model with a simpler model like logistic regression reduces the computational complexity per inference. Gradient boosting involves traversing many decision trees, each requiring multiple conditional checks and arithmetic operations, while logistic regression is a single linear transformation. This directly lowers CPU utilization per request, reducing latency under the same instance resources.

Answer

Use a larger instance type, such as ml.c5.4xlarge

Answer

Switch to batch transform for inference

Answer

Enable automatic scaling on the endpoint

Question 9

Drag and drop the steps to set up Amazon SageMaker Ground Truth for a labeling job in the correct order.

Accepted Answer

Prepare dataset, Create labeling job, Configure task settings, Provide labeling instructions, Execute labeling job. Ground Truth setup involves dataset preparation, job creation, task configuration, instructions, and execution.

Answer

Prepare dataset, Create labeling job, Provide labeling instructions, Configure task settings, Execute labeling job

Answer

Create labeling job, Prepare dataset, Configure task settings, Provide labeling instructions, Execute labeling job

Answer

Prepare dataset, Create labeling job, Configure task settings, Execute labeling job, Provide labeling instructions

Question 10

Drag and drop the steps to use Amazon SageMaker Debugger to debug a training job in the correct order.

Accepted Answer

Configure SageMaker Debugger hooks → Set up training job with Debugger rules → Execute training job → Analyze debug data in SageMaker Studio. Debugger requires hook configuration, job setup with rules, execution, and analysis.

Answer

Configure SageMaker Debugger hooks → Execute training job → Set up training job with Debugger rules → Analyze debug data

Answer

Execute training job → Configure SageMaker Debugger hooks → Set up training job with Debugger rules → Analyze debug data

Answer

Set up training job with Debugger rules → Configure SageMaker Debugger hooks → Execute training job → Analyze debug data

Question 11

Match each ML model evaluation concept to its definition.

Accepted Answer

Precision: The proportion of true positive predictions among all positive predictions. Precision, Recall, F1 Score, and ROC AUC are key evaluation metrics. Common confusions include swapping Accuracy, Specificity, and Precision definitions.

Answer

ROC AUC: The proportion of correct predictions among total predictions

Answer

Accuracy: The ability to correctly identify negative cases

Answer

Specificity: The proportion of true positive predictions among all positive predictions

Question 12

Refer to the exhibit. A training job failed with the error shown. What is the most likely cause?

Accepted Answer

The training data contains missing values or outliers that cause numerical instability. The error message explicitly states that the input contains NaN or infinity, which indicates missing values or outliers in the training data. This causes numerical instability during training. Option A is incorrect because the error is about input data, not model architecture. Option C is incorrect because the error is from the training script, not insufficient memory. Option D is incorrect because the error is about input values, not runtime limits.

Answer

The model architecture is incorrect

Answer

The instance type does not have enough memory

Answer

The training job exceeded the maximum runtime

Question 13

A data engineering team is building a real-time data pipeline using Amazon Kinesis Data Streams with AWS Lambda for processing. The pipeline ingests clickstream data from a mobile app. The team notices that occasionally, a Lambda function fails due to a transient error, and the failed record is not retried, leading to data loss. The Lambda function is configured with a batch size of 100 and a maximum retry count of 0. The team wants to ensure that all records are processed successfully, even if transient failures occur. They also want to minimize the impact of poison pill records that could block processing. Which combination of actions should the team take to address this issue?

Accepted Answer

Set the maximum retry count to 5, configure an on-failure destination Amazon SQS queue, and set up a dead-letter queue on that SQS queue for poison pills.. To address the issue of data loss due to transient errors and poison pill records, the team should increase the Lambda function's maximum retry count to 5 to allow retries on transient failures. However, even with retries, some records may fail repeatedly (poison pills) which can block the shard if not handled. Configuring an on-failure destination (such as an Amazon SQS queue) on the Lambda function sends all records that failed after retries to that queue. Then, by setting up a dead-letter queue on that SQS queue, poison pill records are isolated and can be examined or reprocessed separately, preventing them from blocking the main processing pipeline. Option A is incorrect because a dead-letter queue on Lambda alone is not sufficient – it captures failures after retries if configured, but the key is to also have an on-failure destination to offload failures. Option B is incorrect because Kinesis Data Firehose is designed for streaming data to destinations like S3, not for real-time per-record Lambda processing with built-in retry logic; it would change the architecture. Option D is incorrect because reducing batch size to 1 would increase costs and processing time, and may not fully resolve transient errors or poison pill issues.

Answer

Set the maximum retry count to 5 and configure a dead-letter queue on the Lambda function to capture failed records after retries.

Answer

Switch to using Amazon Kinesis Data Firehose to buffer data and use AWS Lambda for transformation with built-in retry logic.

Answer

Reduce the batch size to 1 and increase the Lambda function timeout to handle transient errors.

Question 14

A data scientist is training a neural network on Amazon SageMaker and wants to automatically stop training if the validation loss does not improve for 5 consecutive epochs. Which feature should they use?

Accepted Answer

Implement early stopping in the training script. Early stopping is a technique where training is halted when a monitored metric, such as validation loss, stops improving for a specified number of epochs (patience). In SageMaker, this is implemented within the training script itself, often using framework callbacks like Keras EarlyStopping or PyTorch's ReduceLROnPlateau with early stopping logic. SageMaker Debugger is used for monitoring and profiling but does not automatically stop training; it can emit alerts but requires custom rules or hooks to trigger stopping. SageMaker Checkpointing saves model state periodically to resume training, not stop it. SageMaker Hyperparameter Tuning launches multiple training jobs to find optimal hyperparameters, not to stop a single job early. Therefore, option A is correct: the data scientist should implement early stopping in the training script.

Answer

SageMaker Debugger

Answer

SageMaker Checkpointing

Answer

SageMaker Hyperparameter Tuning

Question 15

A machine learning team is using Amazon SageMaker to train models on a large dataset stored in Amazon S3. The dataset is 5 TB in size and is partitioned by date. The team wants to minimize data transfer costs and reduce training time by caching frequently accessed data locally on the training instances. The training instances are EC2 instances with attached Amazon EBS volumes. The team is considering using SageMaker Pipe mode to stream data directly from S3, but they are concerned about network bandwidth. Which approach should the team use to optimize data loading for training?

Accepted Answer

Use Amazon FSx for Lustre as a high-performance file system linked to the S3 bucket, and mount it on the training instances.. Amazon FSx for Lustre is natively integrated with Amazon SageMaker as a data source, providing a high-performance file system that can be linked directly to an S3 bucket. It automatically caches frequently accessed data from S3 on the file system, reducing data transfer costs and training time by avoiding repeated downloads. The caching capability addresses network bandwidth concerns effectively. Option B is incorrect: SageMaker File mode uses EBS volumes, not Amazon EFS, and is not designed as a shared, cached file system across training jobs. Option C is incorrect: copying the entire 5 TB dataset to EBS volumes before each training job is time-consuming, increases costs, and does not provide efficient caching across jobs. Option D is incorrect: SageMaker Pipe mode streams data directly from S3 without caching, so it does not reduce repeated data transfers and may still face bandwidth issues.

Answer

Use SageMaker File mode with Amazon EFS, which allows multiple training instances to share the same file system and caches data from S3.

Answer

Increase the size of the EBS volumes attached to the training instances and copy the entire dataset to the volumes before training.

Answer

Use SageMaker Pipe mode to stream data from S3 directly to the training algorithm, which automatically caches data in memory.