AWS Certified Machine Learning Engineer Associate MLA-C01 (MLA-C01) — Questions 376–450

Why this answer

How to eliminate wrong answers

Option B is wrong because SageMaker Model Monitor is designed for detecting data drift and model quality degradation over time, not for comparing hyperparameter configurations across training runs. Option C is wrong because the SageMaker Experiments SDK is a programmatic API for logging and querying experiment data, not an interactive visual interface. Option D is wrong because SageMaker Debugger Insights provides debugging and profiling information during training, such as gradients and tensors, but lacks the interactive experiment comparison capabilities needed for hyperparameter analysis.

Why this answer

How to eliminate wrong answers

Option B is wrong because API Gateway is a front-end service for creating RESTful APIs; it cannot proxy requests directly to a model stored in S3 — you would still need a compute layer (like Lambda) to load the model and run inference. Option C is wrong because downloading the model from S3 on every invocation introduces significant latency and cost, and may cause timeouts or throttling under load; models should be loaded once and reused across invocations. Option D is wrong because mounting an EFS file system adds complexity, cost, and potential cold-start delays, and is overkill for a static pickle file that can be included directly in a layer; EFS is better suited for large, dynamic datasets that need concurrent access across multiple functions.

Why this answer

Why this answer

Why this answer

Why this answer

Why this answer

How to eliminate wrong answers

Option A is wrong because Amazon Kinesis Data Analytics is used for real-time SQL or Apache Flink-based analytics on streaming data, not for reliably ingesting and storing raw data to S3; it lacks built-in delivery to S3. Option B is wrong because AWS Glue ETL jobs are batch-oriented processing tools that run on a schedule or trigger, not designed for real-time streaming ingestion from a website. Option D is wrong because Amazon Kinesis Data Streams is a real-time data streaming service that requires a separate consumer (e.g., Lambda, Kinesis Data Firehose, or custom application) to write to S3, and Amazon SageMaker is a machine learning platform, not a storage destination for raw clickstream data.

Why this answer

Why this answer

Why this answer

How to eliminate wrong answers

Option A is wrong because AWS Glue ETL running in micro-batches incurs higher costs due to the need for a continuously running Spark environment, and it introduces latency that is unnecessary for simple per-record transformations. Option B is wrong because Kinesis Data Analytics for SQL is designed for complex streaming analytics (e.g., aggregations, windowed queries) and adds unnecessary complexity and cost for simple per-record transformations like adding timestamps or filtering. Option D is wrong because using the Kinesis Client Library (KCL) requires you to provision and manage your own compute (e.g., EC2 instances or containers) to run the consumer application, leading to higher operational overhead and cost compared to the fully managed Firehose + Lambda approach.

Why this answer

How to eliminate wrong answers

Option A is wrong because manual versioning in S3 key names lacks built-in model metadata tracking, approval workflows, and deployment integration, making it error-prone and unscalable for a team of five. Option B is wrong because AWS Config is a service for auditing and evaluating resource configurations (e.g., compliance rules), not for versioning or managing ML model artifacts. Option C is wrong because AWS CodeArtifact is a package management service for software libraries (e.g., Python packages, Maven artifacts), not for storing and versioning trained ML model artifacts or their metadata.

Why this answer

How to eliminate wrong answers

Option B is wrong because step scaling with a lower cooldown (60 seconds) does not address the root cause: the scaling action still takes ~3 minutes to launch new instances, and reducing cooldown only affects how quickly subsequent scaling actions can occur, not the initial delay. Option C is wrong because replacing two m5.xlarge instances with one m5.2xlarge instance reduces total compute capacity (from 8 vCPUs to 4 vCPUs), making the endpoint more vulnerable to bursts and increasing the likelihood of 503 errors. Option D is wrong because scheduled scaling adds instances 5 minutes before the expected peak hour, but the burst is unpredictable and occurs within 30 seconds, so scheduled scaling cannot react to sudden, unplanned traffic spikes.

Why this answer

Why this answer

Why this answer

Why this answer

How to eliminate wrong answers

Option A is wrong because data leakage typically causes both training and test accuracy to be artificially high or inconsistent, not a clear gap where training accuracy is high and test accuracy is low. Option B is wrong because feature scaling (e.g., normalization or standardization) does not cause overfitting; it is a preprocessing step that helps gradient descent converge and can improve model performance, but it does not create the described accuracy disparity. Option D is wrong because underfitting results in poor accuracy on both training and test sets, not high training accuracy with low test accuracy.

Why this answer

Why this answer

Why this answer

Why this answer

How to eliminate wrong answers

Option A is wrong because configuring Studio to run in a private subnet and disabling internet access only restricts internet-bound traffic but does not encrypt the data exchanged between the Studio app and the kernel gateway; it lacks the encryption requirement. Option B is wrong because using a VPC endpoint for SageMaker and enabling KMS encryption only encrypts data at rest and traffic through the endpoint, but does not encrypt the inter-container traffic between the Studio app and the kernel gateway, nor does it restrict all Studio traffic to the VPC. Option D is wrong because enabling network isolation and using a VPC with a NAT gateway isolates the container from the internet but does not encrypt inter-container traffic, and a NAT gateway actually allows outbound internet access, contradicting the requirement to restrict access to only the VPC.

Why this answer

Why this answer

Why this answer

Why this answer

How to eliminate wrong answers

Option A is wrong because AWS Database Migration Service (DMS) is designed for migrating databases to AWS, not for ingesting streaming data from IoT devices. Option B is wrong because AWS Glue ETL jobs are batch-oriented and triggered by events (e.g., S3 PUT), not designed to continuously collect and deliver streaming data in near real time. Option D is wrong because Amazon S3 Transfer Acceleration speeds up uploads over long distances using edge locations but does not provide streaming ingestion or buffering capabilities.

Why this answer

How to eliminate wrong answers

Option A is wrong because one-hot encoding 1,200 categories and 50,000 products would create 51,200 binary columns, causing extreme sparsity and memory blowup, which undermines the factorization machine's efficiency and can lead to poor generalization. Option B is wrong because dropping the 'category' column discards valuable hierarchical information (e.g., product type) that could improve recommendation quality; factorization machines are designed to handle high-cardinality features, so there is no need to drop it. Option D is wrong because PCA is a linear dimensionality reduction technique for continuous features, not suitable for categorical data; applying PCA to one-hot encoded categories would destroy the interpretability of interactions and is not a standard preprocessing step for factorization machines.

Why this answer

Why this answer

Why this answer

How to eliminate wrong answers

Option A is wrong because switching to a multi-model endpoint does not reduce costs when a single model is deployed; it shares instances across multiple models, but the underlying instance still incurs hourly charges regardless of utilization. Option B is wrong because reducing the number of instances to one may still leave the single instance idle most of the time, and the cost of that one instance (e.g., ml.t4.medium at ~$0.0468/hour) would still be significantly higher than serverless pay-per-inference pricing for low-traffic scenarios. Option D is wrong because asynchronous inference endpoints are designed for large payloads or long-running inferences, not for reducing costs on low-utilization endpoints; they still use provisioned instances with hourly billing.

Why this answer

How to eliminate wrong answers

Option A is wrong because not shuffling training data affects model training convergence and generalization, but it does not cause accuracy to degrade over time after deployment; it is a one-time training issue. Option B is wrong because compiling a model for inference (e.g., using SageMaker Neo) optimizes latency and throughput, not accuracy; it has no impact on prediction quality degradation. Option D is wrong because an endpoint instance type that is too small would cause performance issues like high latency or throttling, not a gradual decline in model accuracy.

Why this answer

How to eliminate wrong answers

Why this answer

Why this answer

Why this answer

How to eliminate wrong answers

Option A is wrong because the output_path is correctly placed under the MonitoringOutputConfig in the exhibit; SageMaker Model Monitor requires the output location to be specified within MonitoringOutputConfig, not as a separate top-level parameter. Option C is wrong because the MonitoringType should be 'DataQuality' for data quality monitoring, not 'ModelQuality'; 'ModelQuality' is used for model quality monitoring (e.g., accuracy, precision), which is a different monitoring type. Option D is wrong because the cron expression 'cron(0 * * * ? *)' is correctly formatted for an hourly schedule at the start of each hour; there is no syntax error in the expression.

Why this answer

Why this answer

Why this answer

How to eliminate wrong answers

Option A is wrong because AWS RAM is used to share resources like VPC subnets, Transit Gateways, or License Manager configurations, not S3 buckets or SageMaker models; S3 bucket policies or IAM roles are required for cross-account S3 access. Option C is wrong because SageMaker Model Registry does not natively support cross-account sharing; it is a metadata and versioning service within a single account, and sharing models across accounts still requires manual artifact replication or resource policies. Option D is wrong because copying model artifacts to each account's S3 bucket is inefficient, introduces synchronization overhead, and violates the principle of a single source of truth; the recommended approach uses resource-based policies to avoid duplication.

Why this answer

How to eliminate wrong answers

Option A is wrong because the error is an access denied on a read operation, not a write operation; s3:PutObject is irrelevant to reading data. Option B is wrong because the policy does specify the bucket ARN without `/*` for the s3:ListBucket permission (as required), but the issue is that the s3:ListBucket permission itself is missing entirely. Option C is wrong because the Glue job role does have some permissions (as shown in the exhibit), but the specific missing permission is s3:ListBucket, not a general lack of permissions.

Why this answer

How to eliminate wrong answers

Option B (Accuracy) is wrong because accuracy treats false positives and false negatives equally, which is inappropriate when the cost of false negatives is much higher; a model with high accuracy could still miss many positive cases. Option C (F1 score) is wrong because it balances precision and recall, but when false negatives are far more costly, the optimal trade-off should heavily favor recall over precision, not balance them equally. Option D (Precision) is wrong because precision focuses on minimizing false positives, which is the opposite of the requirement; optimizing precision would reduce false alarms but could increase false negatives.

Why this answer

How to eliminate wrong answers

Option B is wrong because improper feature scaling typically leads to slow convergence or suboptimal performance during training, but it would not cause low training and validation errors with poor test performance; scaling issues usually affect both training and validation errors similarly. Option C is wrong because overfitting would result in low training error but high validation error, not low validation error as described in the scenario. Option D is wrong because high bias (underfitting) would cause both training and validation errors to be high, not low.

Why this answer

How to eliminate wrong answers

Option B is wrong because imputing age with the mean is inappropriate when values include '150', '0', and negative numbers, which would skew the mean and introduce bias; also, Standard Scaler should be applied after cleaning and splitting, not during initial prototyping, and DataBrew does not natively support Standard Scaler as a built-in transform. Option C is wrong because replacing age with median still contaminates the dataset with invalid values (e.g., negative numbers) and DataBrew does not support SMOTE (Synthetic Minority Oversampling Technique) as a built-in transform; SMOTE is typically applied in SageMaker or custom scripts. Option D is wrong because dropping the diagnosis_code column removes potentially predictive information without attempting to standardize it, and Random Undersampling would discard 20% of the majority class, which may lead to loss of valuable data and is less preferred than oversampling for a 60/40 imbalance.

Why this answer

How to eliminate wrong answers

Option A is wrong because SageMaker Feature Store is a centralized repository for storing, managing, and sharing features (input data) for ML models, but it does not track the lineage of training steps, hyperparameters, or model artifacts. Option C is wrong because SageMaker Experiments focuses on organizing and comparing multiple training runs (trials) with their parameters and metrics, but it does not automatically capture the full lineage graph connecting datasets, models, and endpoints for audit trails. Option D is wrong because SageMaker Model Registry is a catalog for managing model versions, approvals, and deployments, but it does not track the upstream lineage of how a model was trained (e.g., which dataset version and hyperparameters were used).

Why this answer

How to eliminate wrong answers

Option A is wrong because AWS Step Functions with a scheduled trigger relies on a fixed time interval (e.g., cron expression), not on real-time S3 events, so it cannot react immediately to new data. Option B is wrong because Amazon Simple Workflow Service (SWF) is a legacy workflow orchestration service designed for human-in-the-loop tasks and does not natively integrate with S3 events to trigger SageMaker jobs. Option D is wrong because Amazon Simple Queue Service (SQS) with a polling script requires a separate compute resource (e.g., EC2 or Lambda) to poll the queue and trigger the job, adding latency and complexity compared to EventBridge's direct push-based integration.

Why this answer

Why this answer

How to eliminate wrong answers

Option B is wrong because a cron-based schedule polls for new data on a fixed interval, which is not event-driven; it introduces latency and unnecessary invocations when no new data has arrived, and it does not react immediately to S3 events. Option C is wrong because while S3 event notifications can invoke a Lambda function, this approach bypasses Step Functions orchestration, making it harder to manage complex retraining workflows, error handling, and monitoring model drift as part of a coordinated pipeline. Option D is wrong because SageMaker Pipelines with a schedule trigger is not event-driven; it relies on a time-based trigger rather than reacting to new data arrival in S3, and it lacks the flexibility of Step Functions for integrating with other AWS services for drift monitoring.

Why this answer

Why this answer

Why this answer

Why this answer

How to eliminate wrong answers

Option A is wrong because switching to a larger instance type with more GPU memory may reduce latency if the model is memory-bound, but it is not the most effective first step—it increases cost and does not address software-level inefficiencies. Option C is wrong because SageMaker Model Monitor is used for capturing inference data to detect data drift and model quality issues, not for reducing latency. Option D is wrong because increasing the number of instances (horizontal scaling) improves throughput and handles more concurrent requests, but it does not reduce the latency of individual inference requests; it may even add network overhead.

Why this answer

How to eliminate wrong answers

Option A is wrong because S3 Select operates on a single object at a time and cannot filter across multiple objects or partitions; applying it to 2 billion records would require iterating over all objects, negating cost savings. Option C is wrong because listing all objects in the bucket and reading only those with desired prefixes still requires enumerating the entire bucket, which incurs significant API costs and does not minimize data scanned (the script must still list all objects). Option D is wrong because SageMaker Processing with Input Mode 'File' downloads the entire dataset to the training instance; specifying a prefix for the last 6 months would still download all files under that prefix, but the data is partitioned by year and month, so using the prefix alone does not guarantee partition pruning—the team would need to explicitly list only the relevant prefixes, which is inefficient compared to Glue+Athena.

Why this answer

Why this answer

Why this answer

Why this answer

Why this answer

Why this answer

Why this answer

Why this answer

Why this answer

Why this answer

Why this answer

How to eliminate wrong answers

Option A is wrong because SageMaker Clarify is used for bias detection and explainability of model predictions, not for monitoring input feature distribution drift. Option B is wrong because SageMaker Edge Manager manages and optimizes models on edge devices, focusing on deployment and inference at the edge, not on monitoring input data quality in a cloud-based SageMaker endpoint. Option C is wrong because SageMaker Model Monitor – Model Quality Monitoring tracks prediction quality metrics (e.g., accuracy, precision) against a ground truth, not the distribution of input features.

Why this answer

How to eliminate wrong answers

Option B (AWS CloudFormation) is wrong because it is an Infrastructure as Code (IaC) service for provisioning and managing AWS resources declaratively, not for orchestrating event-driven deployment workflows with conditional logic and error handling. Option C (Amazon EventBridge) is wrong because it is a serverless event bus for routing events between services, but it lacks built-in workflow orchestration capabilities like sequencing, branching, and human approval steps required for deployment pipelines. Option D (AWS CodePipeline) is wrong because it is a CI/CD service focused on source code build, test, and deploy stages, but it does not natively integrate with SageMaker model registry approval workflows or provide the granular orchestration needed for ML model deployment from registry to endpoint.

Why this answer

How to eliminate wrong answers

Option B is wrong because increasing the CPU target to 90% does not address the root cause: CPU utilization never exceeded 70% during the surge, so the policy would still not trigger scaling. Option C is wrong because changing to a larger instance type (ml.c5.4xlarge) provides more memory per instance but does not enable automatic scaling; the endpoint would still be a single instance and could become overwhelmed under similar traffic spikes. Option D is wrong because a scheduled scaling policy assumes predictable event timing, but the question describes a major live event where the timing may be known; however, the team wants a reactive scaling mechanism that triggers automatically before performance degrades, not a pre-scheduled one that may not align with actual traffic patterns.

Why this answer

How to eliminate wrong answers

Option A is wrong because converting to CSV does not address the I/O bottleneck; File mode still downloads the entire dataset to the training instance's local storage before training begins, causing high latency. Option B is wrong because gzip compression reduces file size but File mode with local caching still requires a full download to disk, and decompression adds CPU overhead without eliminating the I/O bottleneck. Option D is wrong because splitting data into smaller files and using multiple instances parallelizes computation but does not reduce per-instance I/O latency; each instance still uses File mode by default, so the bottleneck persists.

Why this answer

Why this answer

How to eliminate wrong answers

Option A is wrong because SageMaker Debugger is used for debugging training jobs (e.g., monitoring gradients, weights, and loss during training), not for monitoring data drift in production inference. Option B is wrong because SageMaker Pipelines is a CI/CD orchestration tool for building and managing ML workflows, not a continuous monitoring service; while AWS Lambda could be used to process drift alerts, the core drift detection capability is missing. Option C is wrong because SageMaker Clarify is designed for bias detection and explainability (SHAP values) on datasets or during training, not for real-time drift monitoring of production endpoints; AWS Config tracks resource configuration changes, not model performance or data drift.

Why this answer

How to eliminate wrong answers

Option A is wrong because 'Encode categorical' transform is used to convert categorical variables into numerical representations (e.g., one-hot encoding), not to parse date strings. Option B is wrong because 'Scale values' transform normalizes or standardizes numerical features (e.g., min-max scaling, z-score), which is irrelevant for extracting date components. Option D is wrong because 'Handle missing' transform addresses null or missing values through imputation or deletion, not date parsing.

Why this answer

Why this answer

How to eliminate wrong answers

Option A is wrong because hash encoding can cause collisions (different categories mapping to the same hash value), which degrades model performance, and using Apache Spark on Amazon EMR adds unnecessary complexity and cost for a task that SageMaker Data Wrangler handles natively. Option B is wrong because one-hot encoding with over 1000 unique values creates over 1000 sparse binary columns, leading to the curse of dimensionality, memory issues, and poor performance in random forests. Option C is wrong because label encoding assigns arbitrary integer values (e.g., 1, 2, 3) that imply ordinal relationships, which random forests can misinterpret as meaningful order, introducing bias and reducing model accuracy.

Why this answer

Why this answer

How to eliminate wrong answers

Option A is wrong because enabling public internet access exposes the Studio environment to the internet, which contradicts the requirement of restricting access to the corporate network; AWS WAF protects against web exploits but does not enforce network-level access control. Option B is wrong because placing SageMaker Studio in a public subnet still allows internet access via an internet gateway, and security groups alone cannot prevent traffic from leaving the VPC or enforce corporate network-only access without additional routing controls. Option D is wrong because IAM policies can restrict API access based on IP addresses but do not control network-level access to the Studio UI or kernel gateway; S3 bucket keys only encrypt data at rest in S3, not all data processed in Studio (e.g., EFS, notebook instances).

Why this answer