AWS Certified Machine Learning Engineer - Associate

Question 1

A data scientist is training a deep neural network on a sparse text-classification dataset using SageMaker. Training with plain stochastic gradient descent (SGD) at a fixed learning rate converges very slowly, and different features clearly need different effective learning rates because some tokens appear frequently while others are rare. The team wants an optimizer that adapts the per-parameter learning rate automatically to speed up convergence without extensive manual learning-rate tuning. Which optimizer change best addresses this need?

Accepted Answer

Switch to the Adam optimizer, which maintains per-parameter adaptive learning rates using running estimates of first and second moments of the gradients.

Answer

Keep SGD but increase the global fixed learning rate by 10x to force faster convergence across all parameters.

Answer

Keep SGD and add L2 weight decay to the loss so that rare-feature weights receive larger updates.

Answer

Switch to plain batch gradient descent computed over the full dataset each step to smooth the gradient signal.

Question 2

A media company deploys a document-processing model that accepts PDF uploads up to 500 MB and can take several minutes per request to run OCR and summarization. Requests arrive in unpredictable bursts, and during idle periods the team wants to avoid paying for running instances. Clients do not need an immediate synchronous response and can poll for results. Which SageMaker deployment option best fits these requirements?

Accepted Answer

SageMaker Asynchronous Inference with autoscaling configured to scale in to zero instances

Answer

A real-time endpoint with target-tracking autoscaling on invocations per instance

Answer

SageMaker Serverless Inference with provisioned concurrency

Answer

A batch transform job triggered on a fixed hourly schedule

Question 3

A data science team stores raw event logs as compressed JSON in Amazon S3. They repeatedly run the same complex SQL aggregation in Amazon Athena to build a feature table for model training, and each run scans hundreds of gigabytes, causing high query costs and slow iteration. They want a repeatable, cost-efficient way to persist the engineered feature dataset in an analytics-friendly format for downstream SageMaker training jobs. Which approach best meets these requirements with the least ongoing operational overhead?

Accepted Answer

Use an Athena CREATE TABLE AS SELECT (CTAS) statement to materialize the aggregated features into a new S3 location as partitioned Parquet, then point training jobs at that output.

Answer

Increase the Athena query result cache retention and rerun the same SELECT query before each training job so results are served from cache.

Answer

Export the query results manually from the Athena console to CSV and upload the file to the training bucket before each run.

Answer

Provision an Amazon Redshift cluster, load the raw JSON, and run the aggregation there for every training iteration.

Question 4

A data scientist trains a binary classifier to predict whether a customer will churn. Only 4% of customers in the dataset actually churn. The initial model reports 96% accuracy on the validation set, but the business complains that almost no real churners are being flagged. Which evaluation approach should the data scientist adopt to properly assess and compare model performance for this problem?

Accepted Answer

Evaluate the model using the area under the precision-recall curve (AUPRC) along with recall for the churn class, because these metrics focus on performance on the rare positive class.

Answer

Continue optimizing for overall accuracy, since 96% indicates the model generalizes well to the validation set.

Answer

Switch to root mean squared error (RMSE) to measure the distance between predicted and actual churn labels.

Answer

Increase the size of the training dataset until accuracy exceeds 99%, which will automatically improve churn detection.

Question 5

A machine learning engineer is training a deep neural network image classifier on SageMaker. Using a very large batch size to maximize GPU throughput, they observe that training loss converges quickly but validation accuracy is noticeably lower than a prior run that used a smaller batch size. Compute budget allows either approach. Which adjustment is MOST likely to improve generalization while keeping training stable?

Accepted Answer

Reduce the batch size (and scale the learning rate accordingly) so the model sees noisier gradient estimates that can improve generalization

Answer

Increase the batch size further to reduce gradient variance and speed convergence

Answer

Remove all regularization and dropout so the model can fit the training data more completely

Answer

Disable shuffling of the training data between epochs to make gradient updates more consistent

Question 6

A retail company retrains a churn-prediction model weekly. Every night, they need to generate churn scores for their entire customer base of 40 million records stored as Parquet files in S3. The scoring job must complete before business hours, but there is no requirement for individual real-time predictions during the day. The ML engineer wants the most cost-effective deployment approach that avoids paying for idle compute. Which SageMaker deployment option should the engineer choose?

Accepted Answer

Use a SageMaker Batch Transform job scheduled nightly to process the S3 dataset

Answer

Deploy a real-time endpoint with autoscaling and invoke it in a loop each night

Answer

Deploy a Serverless Inference endpoint and send each record as an individual invocation

Answer

Configure an Asynchronous Inference endpoint and submit the full dataset as one payload

Question 7

A data scientist trains a binary classifier that flags manufacturing defects. The training set contains 4% defective units. The initial model achieves 96% accuracy but misses most defects during validation. Management requires that the model catch as many true defects as possible while keeping false alarms manageable, and they want a single metric that balances both concerns for model selection. Which evaluation metric should the team optimize?

Accepted Answer

F1 score, the harmonic mean of precision and recall

Answer

Overall classification accuracy on the validation set

Answer

R-squared computed against the defect labels

Answer

Mean absolute error between predicted and actual classes

Question 8

A machine learning team runs a production SageMaker real-time endpoint serving a recommendation model. They need to deploy a new model version but are concerned that a faulty version could cause a spike in inference errors affecting customers. They want SageMaker to automatically shift traffic to the new version gradually, monitor CloudWatch metrics during the shift, and automatically roll back to the previous version if error thresholds are breached. Which approach should they configure?

Accepted Answer

Use a blue/green deployment with canary traffic shifting and CloudWatch alarm-based auto-rollback in the endpoint deployment configuration

Answer

Create a second endpoint with the new model and manually switch the application's endpoint name after testing

Answer

Deploy the new model as a production variant on the same endpoint and manually adjust variant weights over several days

Answer

Use a batch transform job to validate the new model on historical data before replacing the endpoint

Question 9

A financial services team trains a gradient boosting classifier to predict loan default. Downstream systems consume the model's predicted probabilities directly as risk scores to set interest rates, so the numeric probabilities must reflect true likelihoods (e.g., a predicted 0.30 should default about 30% of the time). During evaluation, the model shows strong ranking performance with AUC of 0.88, but the business team reports the predicted probabilities appear systematically too high in the mid-range. Which evaluation step best addresses whether the probabilities can be trusted as risk scores?

Accepted Answer

Plot a calibration (reliability) curve of predicted probability versus observed frequency and apply probability calibration such as isotonic regression or Platt scaling if miscalibrated

Answer

Increase the classification threshold from 0.5 to 0.7 to reduce the number of high-probability predictions

Answer

Re-evaluate using AUC on a larger holdout set, since a higher AUC guarantees well-calibrated probabilities

Answer

Switch the evaluation metric to F1 score to balance precision and recall for the mid-range predictions

Question 10

A machine learning engineer is training a neural network in a SageMaker training job to classify support tickets into 12 mutually exclusive categories. The labels are provided as integer class indices (0 through 11) rather than one-hot encoded vectors. The final layer uses a softmax activation with 12 units. During training, the loss function raises a shape-mismatch error expecting the target to have the same shape as the model output. Which change resolves the error with the least preprocessing effort?

Accepted Answer

Switch the loss function to sparse categorical cross-entropy, which accepts integer class indices directly.

Answer

Change the final layer activation from softmax to sigmoid so it outputs per-class probabilities independently.

Answer

Reduce the number of output units to 1 and treat the task as ordinal regression on the class index.

Answer

Convert the softmax output to a single argmax value before computing standard categorical cross-entropy.

Question 11

An ML engineering team deploys SageMaker real-time endpoints across development, staging, and production accounts. Currently each environment is provisioned manually through the console, causing configuration drift and inconsistent instance types between accounts. The team wants a programmatic, version-controlled way to define the endpoint infrastructure using a familiar general-purpose programming language (TypeScript), reuse the same construct across all three accounts, and integrate the deployment into their existing CI/CD pipeline. Which approach best meets these requirements?

Accepted Answer

Use the AWS CDK to define the model, endpoint configuration, and endpoint as reusable constructs in TypeScript, synthesize CloudFormation templates, and deploy the same stack to each account through the CI/CD pipeline.

Answer

Write a Python script using the boto3 SDK that calls CreateModel, CreateEndpointConfig, and CreateEndpoint, and run it manually in each account when a deployment is needed.

Answer

Author a hand-written CloudFormation JSON template for each environment separately, editing instance types inline per account before uploading each template to the console.

Answer

Use SageMaker Pipelines steps to create the endpoints and rely on pipeline parameters to switch between accounts at runtime.

Question 12

A machine learning platform team wants to define their SageMaker infrastructure (endpoints, endpoint configs, and IAM roles) as code. The team is composed of Python developers who want to use loops, conditionals, and reusable classes to generate the infrastructure definitions, while still deploying through AWS's native provisioning and rollback engine. Which approach best meets these requirements?

Accepted Answer

Author AWS CDK stacks in Python, which synthesize into CloudFormation templates for deployment

Answer

Write raw CloudFormation templates in YAML and manually parameterize them for each resource

Answer

Use the AWS CLI in a shell script to create endpoints imperatively at deploy time

Answer

Use Terraform with the AWS provider to manage all SageMaker resources

Question 13

A security team wants to detect unusual patterns in network traffic logs to flag potential intrusions. The dataset contains millions of unlabeled records with numeric features such as packet count, byte volume, and connection duration. The team has no labeled examples of attacks and needs a scalable SageMaker built-in algorithm that can identify statistical outliers in this high-dimensional data. Which algorithm best fits this requirement?

Accepted Answer

Random Cut Forest (RCF)

Answer

XGBoost with a binary objective

Answer

Linear Learner in classifier mode

Answer

Factorization Machines

Question 14

A retail company wants to group its 2 million customers into distinct segments based on purchasing behavior, browsing patterns, and demographic attributes. The marketing team has no predefined labels for the segments and wants the model to discover natural groupings in the data so they can design targeted campaigns. Which SageMaker built-in algorithm is most appropriate for this task?

Accepted Answer

K-Means clustering

Answer

XGBoost with multiclass objective

Answer

Linear Learner in classification mode

Answer

Factorization Machines

Question 15

A retail company wants to forecast weekly product demand across 5,000 store-item combinations. The dataset includes historical sales, promotional flags, holiday calendars, and price changes. The team needs a single model that can produce probabilistic forecasts (with prediction intervals) for all series simultaneously and learn shared patterns across related items. Which SageMaker built-in algorithm best fits this requirement?

Accepted Answer

SageMaker DeepAR

Answer

SageMaker Linear Learner

Answer

SageMaker XGBoost with a separate model trained per series

Answer

SageMaker Random Cut Forest

Question 16

A retail company wants to build an image classifier that identifies which of 40 product categories appears in a photo. They have only about 3,000 labeled images total (roughly 75 per category) and limited GPU budget. The ML team needs the fastest path to a reasonably accurate model without collecting more labeled data. Which modeling approach best fits this situation?

Accepted Answer

Use transfer learning with SageMaker's built-in image classification algorithm in transfer-learning mode, fine-tuning a pretrained network on the labeled images

Answer

Train a deep convolutional neural network from scratch on the 3,000 images using random weight initialization

Answer

Use the SageMaker K-Means algorithm to cluster the images into 40 groups and label the clusters by category

Answer

Use the SageMaker Linear Learner algorithm on the raw pixel values to classify the 40 categories

Question 17

A data science team at a real estate company needs to predict house sale prices from a tabular dataset containing 40 numeric and categorical features and about 120,000 rows. The relationships between features and price are highly non-linear, and the team wants strong predictive accuracy with minimal feature engineering while training efficiently on SageMaker. Which modeling approach should they choose?

Accepted Answer

Use the SageMaker XGBoost built-in algorithm configured for regression (reg:squarederror).

Answer

Use the SageMaker Linear Learner algorithm in regression mode.

Answer

Use the SageMaker K-Means algorithm to cluster homes and assign each cluster's mean price.

Answer

Use the SageMaker Random Cut Forest algorithm to score each record's price.

Question 18

A media company wants to test five different article headline variants on its homepage to maximize click-through rate. The traffic is high, and business stakeholders want to minimize the number of impressions wasted on clearly underperforming headlines while the test is running, dynamically shifting more traffic toward better-performing variants as evidence accumulates. Which modeling approach best fits this requirement?

Accepted Answer

A multi-armed bandit approach that adaptively allocates more traffic to higher-performing headlines during the experiment

Answer

A fixed 20/20/20/20/20 A/B/n split held constant until a predetermined sample size is reached, then a chi-square test

Answer

A linear regression model predicting click-through rate from headline text length

Answer

A time-series forecasting model to predict daily click volume for each headline

Question 19

A manufacturing company trains an XGBoost binary classifier in SageMaker to detect defective parts on an assembly line. Only 1.5% of parts are defective. The initial model achieves 98.5% accuracy but misses nearly all defective parts, which is unacceptable because undetected defects cause costly recalls. The team wants to improve the model's ability to catch defects while keeping the training pipeline simple and avoiding synthetic data generation. Which approach BEST addresses this problem?

Accepted Answer

Set the XGBoost scale_pos_weight hyperparameter to increase the penalty for misclassifying the minority (defective) class, and evaluate using recall and F1 rather than accuracy.

Answer

Increase the number of boosting rounds (num_round) and lower the learning rate (eta) to let the model train longer on the existing data.

Answer

Continue using accuracy as the objective metric but add more non-defective samples to increase the total training set size.

Answer

Switch the objective to reg:squarederror so the model outputs continuous defect scores instead of class labels.

Question 20

A team manages a SageMaker real-time endpoint entirely through a CloudFormation stack that is deployed identically across dev, staging, and production. During an incident, an on-call engineer manually changed the production endpoint's instance count and instance type through the SageMaker console to handle a traffic surge. Later, the team wants to programmatically detect whether the running production resources still match what the CloudFormation template declares, before running their next stack update. Which approach lets them identify these out-of-band changes with the least custom effort?

Accepted Answer

Run CloudFormation drift detection on the stack and review the resources reported as MODIFIED.

Answer

Enable termination protection on the stack so no manual changes can occur in the future.

Answer

Create an AWS Config custom rule with a Lambda function that queries the endpoint configuration on every change.

Answer

Delete and recreate the stack so the deployed resources match the template exactly.

Question 21

A financial services company must maintain a tamper-evident audit trail of all administrative API calls made against its SageMaker resources (creating endpoints, updating endpoint configs, deleting models) across every AWS Region. The compliance team needs to detect any unauthorized configuration changes and be able to query historical activity months later. Which approach best meets these requirements?

Accepted Answer

Enable a multi-Region AWS CloudTrail trail that logs management events to an S3 bucket with Object Lock enabled, and analyze the logs with Athena.

Answer

Enable SageMaker Model Monitor with a data quality baseline to capture all API calls and store violations in CloudWatch.

Answer

Turn on SageMaker Debugger and configure rules to flag any endpoint configuration changes across Regions.

Answer

Stream SageMaker endpoint invocation metrics to CloudWatch Logs and set metric filter alarms on administrative actions.

Question 22

An ML engineer runs a fraud-detection endpoint monitored by two SageMaker Model Monitor jobs: one for data quality drift and one for model quality (accuracy) degradation. Each job publishes a CloudWatch metric, and separate CloudWatch alarms already trigger an automated retraining pipeline via EventBridge. The team complains that during noisy input spikes, the data-quality alarm fires transient alerts and needlessly launches expensive retraining jobs, even when model accuracy is still acceptable. They want retraining to be triggered only when BOTH drift is detected AND accuracy has actually degraded, without deleting the existing individual alarms used for dashboards. What is the most operationally efficient way to meet this requirement?

Accepted Answer

Create a CloudWatch composite alarm with an ALARM rule that requires both the data-quality drift alarm and the model-quality alarm to be in ALARM state, and route only the composite alarm to the retraining EventBridge rule.

Answer

Increase the evaluation periods and datapoints-to-alarm on the data-quality alarm so it only fires after prolonged drift, and keep it as the sole retraining trigger.

Answer

Delete the two individual alarms and create a single metric-math alarm that averages the drift and accuracy metrics into one combined threshold.

Answer

Add a Lambda function subscribed to both alarm SNS topics that maintains state in DynamoDB and starts retraining only after receiving both notifications within a time window.

Question 23

A company hosts several real-time SageMaker endpoints for internal analytics teams. The ML platform lead notices the monthly inference bill has grown steadily, but many endpoints show very low invocation counts during business hours and none overnight. The lead wants an automated, low-effort way to identify underutilized endpoints so that the team can decommission or consolidate them, while avoiding manual review of each endpoint. Which approach best meets this goal?

Accepted Answer

Publish the SageMaker Invocations, CPUUtilization, and MemoryUtilization metrics to a CloudWatch dashboard and set CloudWatch alarms that notify the team via SNS when Invocations remains near zero over a multi-day period.

Answer

Create a CloudWatch alarm on the InvocationsPerInstance metric with a low static threshold and configure it to terminate the endpoint automatically.

Answer

Enable SageMaker Model Monitor data-quality monitoring on each endpoint to detect when no data is being captured.

Answer

Switch all endpoints to SageMaker Batch Transform so that idle compute is never provisioned.

Question 24

A machine learning engineer operates a real-time SageMaker endpoint serving a recommendation model. Recently, downstream applications have reported intermittent failures. The engineer wants to be automatically notified whenever the endpoint returns a sustained elevated rate of server-side errors (HTTP 5xx) so the on-call team can respond quickly. Which approach requires the least custom instrumentation to meet this requirement?

Accepted Answer

Create a CloudWatch alarm on the built-in Invocation5XXErrors metric emitted by the endpoint in the AWS/SageMaker namespace, with an SNS action to notify the on-call team.

Answer

Enable SageMaker Model Monitor data quality monitoring and configure it to publish a violation whenever 5xx errors exceed a threshold.

Answer

Add custom logging in the inference container to write error counts to CloudWatch Logs, then build a metric filter and alarm on the parsed count.

Answer

Use AWS X-Ray tracing on the endpoint and configure an alarm on trace error rates to trigger an SNS notification.

Question 25

A healthcare company runs a SageMaker real-time endpoint that processes protected health information (PHI). A compliance auditor requires that all model prediction logs and captured inference data be encrypted at rest with a customer-managed key, and that any attempt to disable logging or delete the key be centrally recorded for audit. The ML engineer has already enabled SageMaker Data Capture writing to an S3 bucket. Which combination of actions BEST satisfies these requirements with the least custom code?

Accepted Answer

Configure the Data Capture S3 destination and CloudWatch Logs log group to use a customer-managed AWS KMS key, and enable AWS CloudTrail with a trail that logs management events including KMS and logging configuration changes.

Answer

Encrypt the S3 bucket with SSE-S3, enable CloudWatch Logs default encryption, and write a Lambda function triggered on a schedule to check whether logging is still enabled.

Answer

Enable server-side encryption on the endpoint with an EBS volume key and configure Amazon Macie to scan the captured data bucket for policy violations daily.

Answer

Use SageMaker Model Monitor with a data quality baseline and rely on its built-in encryption and audit dashboard to track key and logging changes.

Question 26

A machine learning engineer manages a real-time SageMaker endpoint serving a recommendation model. Business stakeholders complain that during peak hours the recommendations feel slow, but the endpoint's average latency in CloudWatch looks acceptable. The engineer needs to detect and alert on the tail-latency experienced by the slowest requests so the team can react before customers are affected. Which CloudWatch approach best addresses this requirement?

Accepted Answer

Create a CloudWatch alarm on the p99 percentile statistic of the ModelLatency metric emitted by the endpoint.

Answer

Create a CloudWatch alarm on the Average statistic of the OverheadLatency metric emitted by the endpoint.

Answer

Create a CloudWatch alarm on the SampleCount statistic of the Invocations metric emitted by the endpoint.

Answer

Enable SageMaker Model Monitor data quality baselines to capture latency drift over time.

Question 27

A retail company runs three SageMaker real-time endpoints, each with Model Monitor data-quality schedules that emit violation counts as CloudWatch metrics. The operations team wants a single view where on-call engineers can see current drift violation counts across all three endpoints side by side, without navigating to each endpoint separately, and share it with stakeholders. Which approach requires the least custom code and best meets this requirement?

Accepted Answer

Build a CloudWatch dashboard with metric widgets that plot the Model Monitor violation metrics from all three endpoints together.

Answer

Enable AWS X-Ray tracing on all three endpoints and view the service map to observe drift violations.

Answer

Stream the Model Monitor CloudWatch metrics into Amazon Kinesis Data Firehose and build a QuickSight dashboard from the raw JSON.

Answer

Write a Lambda function that queries each endpoint's monitoring report from S3 every minute and posts an aggregated summary to an SNS topic.

Question 28

A data science team must build a text-classification model to route incoming support tickets into 12 categories. They have only 400 labeled tickets total but millions of unlabeled historical tickets. Management wants the highest accuracy possible while minimizing new labeling effort. Which modeling approach best fits this situation?

Accepted Answer

Fine-tune a pre-trained language model (transfer learning) on the 400 labeled tickets.

Answer

Train a linear classifier from scratch on the 400 labeled tickets using bag-of-words features.

Answer

Use k-means clustering on all tickets and label each cluster manually as one of the 12 categories.

Answer

Discard the labeled data and rely solely on rule-based keyword matching for each category.

Question 29

A data science team trains a new version of a customer churn classifier weekly. Each version is registered as a new model package version in an existing SageMaker Model Registry model package group. Before promoting a new version to production, the team must compare its offline evaluation metrics (AUC, F1) against the currently approved production version and record the decision, keeping a full history of which version was approved and by whom. Which approach best meets these requirements with the least custom development?

Accepted Answer

Store each version's metrics in the ModelMetrics of the model package, query and compare versions in the Model Registry, and update the ModelApprovalStatus to Approved for the chosen version.

Answer

Write each model's metrics to a custom DynamoDB table and build a Lambda function that compares them and copies the winning model artifact to a fixed 'production' S3 prefix.

Answer

Save evaluation metrics as CloudWatch custom metrics and use a CloudWatch dashboard to visually compare AUC across versions before manually redeploying the endpoint.

Answer

Enable SageMaker Model Monitor on the production endpoint and rely on its drift reports to decide whether the newest trained version should replace the current one.

Question 30

A data science team trains two versions of a binary classifier on the same small validation set of 400 records. Version A reports 88.0% accuracy and Version B reports 88.5% accuracy across a single evaluation run. The lead engineer must decide whether Version B is genuinely better before promoting it in the SageMaker Model Registry. What is the most statistically sound next step?

Accepted Answer

Use k-fold cross-validation to obtain a distribution of scores for each version and compare their mean and variance rather than relying on a single point estimate.

Answer

Promote Version B immediately because its accuracy is numerically higher on the validation set.

Answer

Retrain both versions with a larger learning rate to amplify the accuracy difference and make the comparison clearer.

Answer

Switch the evaluation metric from accuracy to training loss and compare the final loss values of both versions.

Question 31

A data scientist trains a binary classifier to detect fraudulent transactions. Fraud represents only 1.5% of all transactions. On the validation set, the model achieves 98.5% accuracy, but the business reports that many fraudulent transactions still slip through undetected. The confusion matrix shows a very high number of false negatives and very few false positives. Which action will BEST address the business concern?

Accepted Answer

Lower the classification threshold to increase the model's recall for the fraud (positive) class, and evaluate using the F1 score or recall rather than accuracy.

Answer

Continue optimizing for overall accuracy, since 98.5% already exceeds typical production targets.

Answer

Raise the classification threshold so the model is more confident before flagging fraud, reducing false negatives.

Answer

Add more features to the majority (non-fraud) class to strengthen the model's ability to identify legitimate transactions.

Question 32

A data scientist is training a linear model with gradient descent on a dataset where one feature ranges from 0 to 1 and another ranges from 0 to 1,000,000. During training, the loss oscillates and takes many epochs to converge, and reducing the learning rate slows progress to a crawl without fixing the instability. What is the MOST likely cause and the appropriate remedy?

Accepted Answer

The features are on vastly different scales, causing an elongated, ill-conditioned loss surface; standardize or normalize the features before training.

Answer

The model is overfitting; add L2 regularization and reduce the number of training epochs.

Answer

The training data has too few samples; collect more data before retraining the model.

Answer

The model is too simple for the data; switch to a deep neural network with additional hidden layers.

Question 33

A data scientist is building a document deduplication system. Each document is represented as a high-dimensional TF-IDF embedding vector, but document lengths vary widely, so longer documents produce vectors with larger magnitudes. The team wants a similarity metric for a k-nearest-neighbors model that focuses on the orientation (topical content) of vectors rather than their magnitude, so that a short and a long document about the same topic are still judged similar. Which distance/similarity measure best fits this requirement?

Accepted Answer

Cosine similarity, which measures the angle between vectors and is invariant to their magnitude

Answer

Euclidean (L2) distance, which measures straight-line distance in the feature space

Answer

Manhattan (L1) distance, which sums absolute coordinate differences

Answer

Squared Euclidean distance, which emphasizes larger coordinate differences

Question 34

A machine learning engineer is building a neural network image classifier in SageMaker that must output calibrated class probabilities so a downstream business rule can act only when confidence exceeds 90%. During training, the engineer needs to choose a loss function for the final softmax output layer. Which loss function best supports producing well-calibrated probability estimates for this multiclass problem?

Accepted Answer

Categorical cross-entropy loss

Answer

Hinge loss

Answer

Mean squared error loss

Answer

Huber loss

Question 35

A financial services company trains models in the us-east-1 account and copies the resulting model artifacts to an S3 bucket in a disaster-recovery account located in us-west-2. Both buckets use SSE-KMS with customer managed keys. When a SageMaker training job in us-east-1 tries to write artifacts that are then replicated, and when the DR account's SageMaker attempts to deploy the model, the team encounters AccessDenied errors on decryption. What is the MOST likely root cause and correct remediation?

Accepted Answer

The DR account's KMS key policy and the SageMaker execution role in that account do not grant kms:Decrypt on the DR-region key; update the key policy and role permissions to allow the DR SageMaker role to decrypt the replicated objects.

Answer

S3 replication cannot copy SSE-KMS encrypted objects at all; disable encryption on the source bucket so objects replicate in plaintext.

Answer

KMS keys are global resources, so the same key ARN works in both regions; the error is caused by an expired IAM session and will resolve on retry.

Answer

SageMaker automatically re-encrypts artifacts with an AWS managed key on deployment, so the customer managed key permissions are irrelevant; switch both buckets to SSE-S3.

Question 36

A healthcare analytics team ingests raw CSV files into an S3 bucket that will be used to build a training dataset for a readmission-prediction model. Before the data can be used, compliance requires that any personally identifiable information (PII) such as names, phone numbers, and Social Security numbers be automatically discovered and redacted at scale during the ETL process. The team already uses AWS Glue for their transformation jobs. Which approach best meets this requirement with the least custom development?

Accepted Answer

Use the AWS Glue PII detection transform (Detect Sensitive Data) within the Glue ETL job to identify and mask or redact PII columns and values during processing.

Answer

Write a custom PySpark UDF in the Glue job that applies regular expressions to every column to find and remove PII patterns.

Answer

Enable S3 default encryption with SSE-KMS on the bucket so that PII data is protected before the Glue job reads it.

Answer

Run an Amazon Macie classification job on the bucket and rely on Macie to redact the sensitive values in place before Glue reads the files.

Question 37

A data scientist trains an XGBoost model in a SageMaker training job to predict customer churn. After training, the model achieves 98% accuracy on the training set but only 71% accuracy on the validation set. The validation loss begins increasing after several boosting rounds while training loss keeps decreasing. Which action is MOST likely to improve the model's generalization?

Accepted Answer

Apply regularization by increasing the lambda and alpha hyperparameters and reduce max_depth, while using early stopping on the validation metric.

Answer

Increase the max_depth hyperparameter and add more boosting rounds to help the model learn the training data more completely.

Answer

Switch the objective function from binary:logistic to reg:squarederror to better fit the target distribution.

Answer

Collect additional features and increase the number of columns used per split with a higher colsample_bytree value.

Question 38

A machine learning team has trained a large transformer-based text classifier that achieves 94% accuracy on the validation set. When deployed to a real-time SageMaker endpoint, the model exceeds the required p99 latency budget and cannot be served on the available CPU instance family within cost constraints. The team wants to significantly reduce inference latency while retaining as much of the original accuracy as possible, and they are willing to run an additional training phase. Which modeling approach best meets these goals?

Accepted Answer

Apply knowledge distillation to train a smaller student model that learns from the large teacher model's outputs

Answer

Increase the endpoint instance count with target-tracking autoscaling to distribute the load across more nodes

Answer

Enable data capture on the endpoint and use Model Monitor to identify features driving latency

Answer

Retrain the original large model with a higher learning rate to speed up convergence during training

Question 39

A data scientist is training a deep feed-forward neural network on a tabular dataset of 50,000 records to predict customer churn. During training, the model reaches 98% accuracy on the training set but only 74% accuracy on the validation set, and the gap widens as training continues. The team wants to improve generalization without collecting more data. Which change is MOST likely to reduce this generalization gap?

Accepted Answer

Add dropout layers and apply L2 weight regularization to the network

Answer

Increase the number of hidden layers and neurons to give the model more capacity

Answer

Raise the learning rate so the optimizer converges to the training minimum faster

Answer

Train for more epochs until the training accuracy reaches 100%

Question 40

A data scientist is training an XGBoost model in SageMaker to predict loan defaults. They set num_round to 2000 but notice that the validation AUC peaks around round 350 and then slowly degrades while training AUC keeps improving. They want to automatically stop training when the model stops improving on the validation set and keep the best-performing model, without manually rerunning experiments to find the ideal number of rounds. Which approach best achieves this?

Accepted Answer

Set the early_stopping_rounds hyperparameter and provide a validation channel, so training halts if the evaluation metric does not improve for the specified number of consecutive rounds.

Answer

Reduce num_round to 350 based on the observed peak and retrain the model with only a training channel.

Answer

Enable SageMaker automatic model tuning with num_round as the only tunable hyperparameter across 2000 discrete values.

Answer

Increase the learning rate (eta) so the model converges before round 350 and stops overfitting.

Question 41

A data engineer is preparing a customer transactions dataset for a fraud-detection model. One column, 'merchant_id', contains over 50,000 unique values and is highly predictive. The team wants to convert this feature into a numeric representation without creating tens of thousands of sparse columns that would explode dimensionality and slow training. Which feature-engineering approach best addresses this requirement?

Accepted Answer

Use target (mean) encoding to replace each 'merchant_id' with the fraud rate associated with that merchant, computed on the training folds

Answer

Apply one-hot encoding to 'merchant_id' so each unique merchant becomes its own binary column

Answer

Drop the 'merchant_id' column entirely since high-cardinality features cannot be used in machine learning

Answer

Convert 'merchant_id' to a floating-point number by casting the raw identifier string to its numeric value

Question 42

A data scientist has trained a single deep decision tree on a tabular dataset. The model achieves near-perfect accuracy on the training set but performs poorly and inconsistently on the validation set, and the results vary widely when the tree is retrained on slightly different data samples. The scientist wants to keep using tree-based models but reduce this instability without significantly increasing bias. Which ensemble approach is the MOST appropriate choice?

Accepted Answer

Use bagging (e.g., a random forest) to train many trees on bootstrap samples and average their predictions

Answer

Use boosting (e.g., AdaBoost) to sequentially train shallow trees that correct prior errors

Answer

Increase the maximum depth of the single tree to capture more patterns

Answer

Switch to a single linear regression model to reduce complexity

Question 43

A machine learning team at a fraud-detection company uses Amazon SageMaker Feature Store. They need to serve engineered features with single-digit millisecond latency to a real-time inference endpoint that scores credit card transactions as they occur. Separately, a data science team needs to query the full history of those same features to build training datasets and run batch analytics with Amazon Athena. Which configuration best meets both requirements?

Accepted Answer

Create the feature group with both the online store and the offline store enabled, serving low-latency lookups from the online store and running historical queries against the offline store in S3.

Answer

Create the feature group with only the offline store enabled and query it directly from the inference endpoint for each transaction.

Answer

Create the feature group with only the online store enabled and use Amazon Athena to query the online store for historical training datasets.

Answer

Store all features in a single Amazon RDS instance and add a read replica to serve the real-time inference endpoint.

Question 44

A data engineer is preparing a large tabular dataset stored as CSV files in Amazon S3 to train a SageMaker built-in algorithm (Linear Learner). Training jobs are running slowly because the algorithm must parse the raw CSV text on every epoch. The engineer wants to use an AWS Glue ETL job to output the data in a format that SageMaker built-in algorithms can stream efficiently with Pipe mode. Which output format should the Glue job produce?

Accepted Answer

RecordIO-protobuf format

Answer

JSON Lines format

Answer

Gzip-compressed CSV format

Answer

Apache Avro format

Question 45

A data engineering team stores raw ML training data across three S3 prefixes, each containing CSV files with different but overlapping schemas. Analysts need to run ad hoc Athena queries to explore and validate the data before feature engineering, and the team wants to minimize the manual effort of maintaining table definitions as new files arrive daily. Which approach best meets these requirements?

Accepted Answer

Configure AWS Glue crawlers to crawl each S3 prefix on a daily schedule, populating the Glue Data Catalog with table definitions that Athena can query.

Answer

Manually create AWS Glue Data Catalog table definitions for each prefix and update them with a Lambda function whenever new files are uploaded.

Answer

Load all CSV files into an Amazon Redshift cluster and run exploratory queries directly against Redshift tables.

Answer

Use Amazon SageMaker Data Wrangler to import each prefix individually and export inferred schemas to a JSON file for reference.

Question 46

A data scientist is preparing a dataset for a regression model that predicts customer lifetime value. One feature, 'annual_purchase_amount', is heavily right-skewed with a long tail of high-value customers, and the model's linear assumptions are being violated. The team uses AWS Glue DataBrew for their preparation workflow and wants to reduce skewness so the feature approximates a more normal distribution before training. Which transformation should they apply?

Accepted Answer

Apply a logarithmic transformation to the 'annual_purchase_amount' column

Answer

Apply min-max normalization to rescale the values between 0 and 1

Answer

One-hot encode the 'annual_purchase_amount' column into binary indicators

Answer

Replace all values above the 95th percentile with the column mean

Question 47

A data engineering team receives raw event data as JSON files in Amazon S3. Each record contains a nested array field called 'items', where each element holds product attributes for a single purchase. The ML team needs a flat, tabular dataset in which every product within an order becomes its own row so features can be computed per product. The team wants a low-code, visual approach that produces a reusable transformation recipe. Which approach best meets these requirements?

Accepted Answer

Use AWS Glue DataBrew and apply an 'unnest'/'explode' transformation on the 'items' array to flatten nested elements into individual rows within a saved recipe.

Answer

Configure an AWS Glue crawler to detect the schema, then rely on the crawler to automatically flatten the nested array during cataloging.

Answer

Use Amazon Athena to run a SELECT query on the JSON files and export the result set, treating the array column as a single string.

Answer

Load the JSON into SageMaker Feature Store, which will automatically normalize nested arrays into separate feature records.

Question 48

A data engineer is preparing a customer churn dataset for training a logistic regression model in Amazon SageMaker. The dataset contains a 'contract_type' column with three unordered values: 'month-to-month', 'one-year', and 'two-year'. The engineer wants to convert this nominal categorical feature into a numeric representation that does not impose any artificial ordinal relationship between the categories, using a low-code approach. Which transformation should the engineer apply?

Accepted Answer

Apply one-hot encoding in AWS Glue DataBrew to create a separate binary column for each contract type.

Answer

Apply ordinal (label) encoding to map 'month-to-month' to 0, 'one-year' to 1, and 'two-year' to 2.

Answer

Apply min-max scaling to the 'contract_type' column to normalize the values between 0 and 1.

Answer

Hash the 'contract_type' values into a fixed 256-dimension feature vector using feature hashing.

Question 49

A data engineer is preparing a tabular dataset of manufacturing sensor readings for a regression model. Exploratory analysis shows that a small number of temperature readings are physically impossible (e.g., 9999 degrees) due to sensor malfunctions, and these extreme values are skewing the model. The team wants a low-code, visual way to profile the data, identify these anomalies, and apply cleaning transformations before writing the results back to Amazon S3 for training. Which approach best meets these requirements with the least custom code?

Accepted Answer

Use AWS Glue DataBrew to run a data profile job that surfaces outliers and column statistics, then build a DataBrew recipe with transformations to filter or replace the invalid values, and publish the cleaned dataset to S3.

Answer

Write a custom PySpark AWS Glue ETL script that computes z-scores for each numeric column and drops rows exceeding a hardcoded threshold before writing to S3.

Answer

Enable a SageMaker Ground Truth labeling job so human reviewers can flag each impossible temperature reading individually before training.

Answer

Configure an AWS Glue crawler to scan the data and automatically remove statistical outliers based on the inferred schema.

Question 50

A data engineer is preparing a customer transactions dataset in AWS Glue DataBrew for an ML churn model. The 'signup_date' column contains values in several inconsistent string formats (e.g., '01/15/2023', '2023-01-15', 'Jan 15 2023') because the data was merged from three legacy systems. Before the dates can be used to engineer tenure features, they must be standardized into a single consistent date type. What is the most efficient way to accomplish this within the data preparation workflow?

Accepted Answer

Apply a DataBrew date-format transformation that parses the source patterns and outputs a single standardized date format for the column.

Answer

Drop all rows where the 'signup_date' does not already match the ISO 8601 format to guarantee consistency.

Answer

Export the raw column to Amazon S3 and use Amazon Macie to reformat the date strings before re-importing.

Answer

Leave the dates as strings and rely on the SageMaker training algorithm to automatically infer and normalize the formats.

Question 51

A data engineering team uses AWS Glue DataBrew to interactively clean and transform a customer dataset for a churn-prediction model. After building and validating a sequence of cleaning steps in a DataBrew project, they now need to apply the exact same transformations automatically to new monthly data files that arrive in Amazon S3, without repeating the interactive work each month. What is the MOST appropriate way to operationalize these transformations?

Accepted Answer

Publish the project's recipe and create a DataBrew recipe job that runs against the new S3 dataset, scheduling it to run monthly.

Answer

Export the DataBrew project as a Python script and run it manually in a SageMaker notebook each month.

Answer

Create a DataBrew profile job on the new data each month to reproduce the transformations.

Answer

Recreate the interactive DataBrew project each month and re-enter the transformation steps on the new files.

Question 52

A data engineer is preparing a tabular dataset for a k-nearest neighbors model. The dataset includes 'annual_income' (values from 20,000 to 500,000) and 'age' (values from 18 to 90). During evaluation, the team notices that predictions are almost entirely driven by income differences, while age has negligible influence, even though domain experts consider age highly relevant. What is the MOST appropriate feature preparation step to address this?

Accepted Answer

Apply min-max normalization to both 'annual_income' and 'age' so all features fall within the same 0-1 range before training.

Answer

One-hot encode 'age' into discrete buckets so it contributes equally to the distance calculation.

Answer

Drop 'annual_income' from the dataset because its large magnitude introduces bias into the model.

Answer

Apply a log transformation to 'annual_income' only, leaving 'age' unchanged.

Question 53

A data engineer is preparing a tabular dataset of 5 million customer transactions stored as Parquet in Amazon S3 for a binary classification model. The team wants reproducible train/validation/test splits so that the exact same records land in each partition every time the preparation pipeline runs, regardless of the order in which Glue reads the files. Which approach best guarantees deterministic, reproducible splits?

Accepted Answer

Apply a hash function to a stable unique key column (e.g., transaction_id) and assign records to splits based on hash-bucket ranges.

Answer

Use Spark's randomSplit with default settings inside the Glue job so each run produces a fresh random partition of the data.

Answer

Sort the DataFrame by ingestion timestamp and take the first 70% for training, next 15% for validation, and last 15% for testing.

Answer

Let SageMaker automatically shuffle and split the data at training time using its built-in validation channel.

Question 54

A data engineer is preparing a customer dataset for a churn prediction model. The 'subscription_tier' column contains inconsistent values entered by different systems, such as 'Premium', 'premium', 'PREMIUM ', and 'Prem.'. These should all represent the same category before one-hot encoding. The engineer wants a low-code, visual approach to detect and consolidate these variants without writing custom Spark code. Which AWS Glue DataBrew capability best addresses this requirement?

Accepted Answer

Apply a recipe combining a lowercase transform, whitespace trimming, and a categorical mapping (replace) step to standardize the variant values to a single canonical value.

Answer

Use the DataBrew outlier detection transform to remove the inconsistent categorical values from the dataset.

Answer

Run a Glue FindMatches ML transform to deduplicate the rows before encoding.

Answer

Apply quantile binning to group the categorical values into numeric buckets.

Question 55

A data engineer is preparing a ride-sharing demand forecasting dataset in an AWS Glue ETL job. The raw data contains a single 'pickup_timestamp' column. The team wants the model to learn recurring daily and weekly demand patterns (for example, that 11 PM and 1 AM are close in time, and that Sunday and Monday are adjacent). Which feature engineering approach should the engineer apply to the extracted hour-of-day and day-of-week values?

Accepted Answer

Apply sine and cosine transformations to the hour-of-day and day-of-week values to encode their cyclical nature.

Answer

One-hot encode the hour-of-day and day-of-week values into separate binary columns.

Answer

Keep hour-of-day and day-of-week as raw integer values and apply standard (z-score) scaling.

Answer

Convert the timestamp to Unix epoch seconds and use that single numeric column as the feature.

Question 56

A retail company is preparing a customer dataset in AWS Glue for a churn prediction model. The source data was merged from multiple CRM systems and contains many duplicate customer records that are not exact matches — names have typos, addresses use different abbreviations, and phone formats vary. These fuzzy duplicates would inflate certain customers' behavior signals and bias the training data. The data engineering team needs a scalable way to identify and remove these near-duplicate records within their existing Glue ETL workflow. Which approach best addresses this requirement?

Accepted Answer

Use the AWS Glue FindMatches ML transform, label a sample of matching and non-matching record pairs, train the transform, and apply it in the Glue job to identify duplicates.

Answer

Apply the Glue DynamicFrame dropDuplicates method to remove records that are exact duplicates across all columns.

Answer

Write a Spark SQL GROUP BY on the customer email column and keep the first record of each group.

Answer

Enable a Glue crawler with schema deduplication so overlapping records are collapsed automatically during cataloging.

Question 57

A machine learning engineer is preparing a dataset stored in Amazon S3 to train a binary classification model that predicts loan default. Before training, the team wants to measure whether the training data over-represents or under-represents specific demographic groups (such as age bands) relative to the positive label, so they can document and mitigate potential bias. The team wants a managed capability that produces bias metrics like class imbalance and difference in positive proportions in labels with minimal custom code. Which approach best meets this requirement?

Accepted Answer

Run a SageMaker Clarify pre-training bias analysis job that computes bias metrics such as Class Imbalance (CI) and Difference in Proportions of Labels (DPL) across the specified facets.

Answer

Enable Amazon SageMaker Model Monitor data quality baseline jobs to detect demographic imbalance in the training dataset.

Answer

Use AWS Glue DataBrew to apply one-hot encoding on the demographic columns, which automatically rebalances the classes before training.

Answer

Configure SageMaker Ground Truth to relabel the minority-group records so the positive and negative labels are evenly distributed.

Question 58

A data engineer builds an AWS Glue ETL job that reads semi-structured JSON order records from S3 and writes them to a curated Parquet dataset used to train a demand-forecasting model. Some records store the field 'quantity' as an integer (e.g., 5) while others store it as a string (e.g., "5"). When the job writes to Parquet, Glue creates two separate columns, 'quantity_int' and 'quantity_string', causing the downstream training job to fail. Which approach resolves this schema conflict most efficiently within the Glue job?

Accepted Answer

Apply the ResolveChoice transform with the 'cast' action to force the 'quantity' field to a single numeric type before writing to Parquet.

Answer

Enable Glue job bookmarks so that only new records with a consistent schema are processed on each run.

Answer

Re-run the Glue crawler with schema evolution enabled to merge the conflicting columns automatically.

Answer

Drop the 'quantity' column using the DropFields transform and recompute it during model training.

Question 59

A data engineering team runs an AWS Glue Spark ETL job that transforms 3 years of raw transaction logs into a curated dataset used to build ML training sets. Data scientists query this dataset in Amazon Athena, almost always filtering by a specific transaction_date range and by region. Currently the Glue job writes all output as a single set of Parquet files to one S3 prefix, and Athena queries scan the entire dataset, driving up cost and query time. Which change to the Glue job most effectively reduces the data scanned per query?

Accepted Answer

Configure the Glue job to write output partitioned by transaction_date and region using partitionKeys, so Athena can prune partitions when filtering.

Answer

Increase the number of Glue worker nodes (DPUs) so the job produces more, smaller Parquet files across the same single prefix.

Answer

Change the output format from Parquet to CSV so Athena can read rows sequentially and stop scanning earlier.

Answer

Enable Glue job bookmarks so only newly arrived records are processed on each run.

Question 60

A data engineering team ingests semi-structured JSON event logs into an S3 bucket daily. The schema of these events evolves frequently — new optional fields appear and some fields change from single values to arrays. The team uses an AWS Glue crawler to populate a Data Catalog table that a downstream Glue ETL job reads to build ML training features. Recently, the crawler has been creating multiple separate tables and marking fields as incompatible, breaking the ETL job. What is the most appropriate way to handle this evolving schema so the ETL job continues to run reliably?

Accepted Answer

Configure the crawler to update the existing table and choose the option to merge new columns into the existing schema, while using Glue DynamicFrames with ResolveChoice in the ETL job to reconcile ambiguous field types.

Answer

Delete the Data Catalog table before each crawler run so a fresh schema is inferred every time from the latest data.

Answer

Disable the crawler and manually define a fixed schema in the Data Catalog that never changes, forcing all incoming data to conform.

Answer

Convert all incoming JSON to CSV before crawling so that Glue always infers a flat, consistent schema.

Question 61

A data engineer is preparing a dataset in AWS Glue for a fraud-detection model. One numeric feature, transaction_amount, ranges from $1 to $500,000 and is heavily right-skewed, with most values clustered under $200 but a long tail of very large transactions. The team wants to reduce the disproportionate influence of the extreme values on a downstream logistic regression model while preserving the ordering of the data. Which transformation should the engineer apply to this feature?

Accepted Answer

Apply a log transformation to compress the range and reduce skew before scaling

Answer

Apply one-hot encoding to convert transaction_amount into categorical bins

Answer

Drop all transactions above the 95th percentile to remove the long tail

Answer

Replace the extreme values with the column mean using mean imputation

Question 62

A data engineering team runs a nightly AWS Glue ETL job that reads raw clickstream files landing in an S3 bucket, transforms them, and writes the results to a curated S3 location that feeds a daily model retraining pipeline. New files arrive throughout each day, and the team notices that every nightly run reprocesses the entire history of files, causing job duration and cost to grow continuously. They need the job to process only newly arrived data on each run without manually tracking which files were already handled. What is the most operationally efficient way to achieve this?

Accepted Answer

Enable Glue job bookmarks so the job persists state about previously processed data and processes only new files on subsequent runs.

Answer

Add a filter transform in the Glue script that keeps only records with a timestamp greater than the current date.

Answer

Configure an S3 lifecycle policy to move processed files to Glacier so the Glue job no longer reads them.

Answer

Increase the number of Glue DPUs so the full reprocessing completes within the nightly window despite growing data.

Question 63

A data engineering team is building a training dataset for a fraud detection model. Transaction records arrive continuously in an Amazon Kinesis Data Stream, while customer profile data is stored as a nightly export in Amazon S3. The team needs to enrich each transaction with the corresponding customer attributes and land the combined records in S3 for downstream SageMaker training. They want a managed, serverless approach that minimizes operational overhead. Which solution best meets these requirements?

Accepted Answer

Use an AWS Glue streaming ETL job that consumes the Kinesis stream and joins each micro-batch with the customer profile data loaded from S3, then writes the enriched output to S3.

Answer

Configure a Kinesis Data Firehose delivery stream to write raw transactions to S3, then run a manual Amazon Athena CREATE TABLE AS SELECT query each night to join with profiles.

Answer

Launch a persistent Amazon EMR cluster running Spark Structured Streaming to read from Kinesis and perform the join against profile data cached in HDFS.

Answer

Use an AWS Glue batch ETL job triggered once per day that reads the entire Kinesis stream from the beginning and joins it with the S3 profile export.

Question 64

A data engineering team ingests clickstream events into an Amazon Kinesis Data Stream that feeds a downstream ML feature pipeline. Producers are written by multiple teams and occasionally add or rename fields, which has repeatedly broken the Apache Flink consumer that parses the records. The ML team needs to enforce that only backward-compatible schema changes are accepted before records enter the stream, while keeping serialization compact for high-throughput ingestion. What should they implement?

Accepted Answer

Register an Avro schema in the AWS Glue Schema Registry and configure producers to serialize records with compatibility set to BACKWARD, rejecting incompatible schema versions.

Answer

Convert all records to CSV format and add a header row so the consumer can dynamically detect new columns as they appear.

Answer

Run an AWS Glue crawler on the Kinesis stream every five minutes to update the Data Catalog and propagate schema changes to consumers.

Answer

Store each event as a separate JSON object in Amazon S3 and use a Lambda function to validate fields against a hard-coded list before processing.

Exam domain	Exam weight	Practice
Data Preparation for Machine Learning	28%	Practice this topic
ML Model Development	26%	Practice this topic
Deployment and Orchestration of ML Workflows	22%	Practice this topic
ML Solution Monitoring, Maintenance, and Security	24%	Practice this topic

AWS Certified Machine Learning Engineer - Associate (MLA-C01) practice exam & study guide

Start studying MLA-C01

All MLA-C01 study resources

MLA-C01 exam domains

Sample MLA-C01 questions

Key MLA-C01 terms

MLA-C01 frequently asked questions

Data Preparation for Machine Learning (28%)

ML Model Development (26%)

Deployment and Orchestration of ML Workflows (22%)

ML Solution Monitoring, Maintenance, and Security (24%)