The AWS Certified Machine Learning Engineer – Associate (MLA-C01) exam validates your skills in building, training, and deploying machine learning models on AWS. It is ideal for those looking to specialize in machine learning.
What happens when you deploy a model to AWS SageMaker with an incorrect input data schema?
AModel will train correctly
BPrediction will not be returned
CModel will ignore the incorrect schema
DDeployment will fail
Explanation
Incorrect input data schema prevents predictions from being returned, while other options misrepresent the deployment behavior.
Q112
A company needs a way to improve its model performance while avoiding overfitting. Which technique should they use?
ACross-validation
BData augmentation
CFeature scaling
DIncreasing training data size
Explanation
Cross-validation helps assess model performance on different subsets, catching overfitting early. The other options improve training but won't directly prevent overfitting.
Q113
You are configuring a Batch Inference job in Amazon SageMaker. What happens when you select 'Asynchronous'?
AInference happens instantly.
BJob runs and completes in the foreground.
CRemote service waits for your input.
DJob runs independently and returns an output later.
Explanation
Asynchronous jobs run independently, allowing you to continue working. The other options suggest immediate feedback or synchronous processing.
Q114
Which AWS service should be used to deploy machine learning models as REST APIs?
AAmazon S3
BAWS Lambda
CAmazon SageMaker
DAmazon EC2
Explanation
Amazon SageMaker simplifies deploying ML models as REST APIs. The other services can host applications but aren't specifically designed for machine learning inference in this context.
Q115
Which service is best suited for model training on large datasets in AWS?
AAmazon SageMaker
BAWS Lambda
CAmazon EC2
DAmazon RDS
Explanation
Amazon SageMaker is designed for model training, while the others are not specialized for this purpose.
Q116
A company needs to make predictions in real-time using a machine learning model. What AWS service should they choose?
AAmazon Redshift
BAWS Batch
CAmazon SageMaker Endpoint
DAWS Glue
Explanation
Amazon SageMaker Endpoint offers real-time predictions, while the others focus on batch processing or ETL.
Q117
You are configuring an AWS IAM policy to restrict access to S3 buckets. What happens if you set a condition that denies access based on the IP address?
AAccess is always denied.
BAccess is denied only from specified IPs.
CAccess is granted to specified IPs.
DAccess is fully unrestricted.
Explanation
Access is denied to any requests from the specified IPs, while allowing otherwise.
Q118
Which service provides managed machine learning model deployment?
AAmazon SageMaker
BAWS Lambda
CAmazon EC2
DAWS Batch
Explanation
Amazon SageMaker is specifically designed for model deployment, while the others serve different functions.
Q119
A company needs real-time data processing for streaming applications. Which service should they choose?
AAmazon S3
BAmazon Kinesis
CAWS Snowball
DAmazon RDS
Explanation
Amazon Kinesis is tailored for real-time processing, unlike the other options.
Q120
What happens when a model in Amazon SageMaker fails to train properly?
ATraining is retried automatically
BAn error report is generated
CModel outputs previous predictions
DTraining is paused indefinitely
Explanation
An error report is generated to help identify the issue, while the other options are incorrect outcomes.