MLS-C01 Practice Exam 55 — 20 Questions

Question 1

A data scientist is building a text classification model. The dataset contains 10,000 documents, each labeled with one of 5 categories. Which algorithm is most suitable for this task?

Accepted Answer

Naive Bayes. Naive Bayes is highly suitable for text classification because it models the probability of each category given the document's word features using Bayes' theorem with a strong independence assumption. It performs well on high-dimensional sparse data like bag-of-words or TF-IDF representations, and it is particularly effective when the number of documents (10,000) is moderate relative to the vocabulary size, as it requires relatively little training data to estimate parameters.

Answer

Principal Component Analysis (PCA)

Answer

Linear regression

Answer

k-means clustering

Question 2

A machine learning engineer is deploying a model to Amazon SageMaker for real-time inference. The model requires low latency and must handle variable traffic patterns. Which SageMaker feature should the engineer use to automatically scale the number of instances based on demand?

Accepted Answer

SageMaker automatic scaling. SageMaker automatic scaling (Application Auto Scaling) is the correct feature because it allows the engineer to define scaling policies (e.g., based on CPU utilization or request latency) that automatically adjust the number of instances behind a SageMaker endpoint in response to real-time traffic patterns. This ensures low latency by maintaining sufficient capacity during spikes and reducing costs during lulls, without manual intervention.

Answer

Amazon EC2 Auto Scaling

Answer

Elastic Inference

Answer

SageMaker Batch Transform

Question 3

Which THREE techniques are commonly used for feature engineering in exploratory data analysis? (Select THREE.)

Accepted Answer

Extracting date/time components like day of week or hour.. Extracting date/time components such as day of week, hour, or month from a timestamp is a standard feature engineering technique. It transforms a single datetime column into multiple categorical or cyclical features that can reveal temporal patterns like weekly seasonality or peak hours, which are often critical for time-series models.

Answer

Using principal component analysis (PCA) to create new features.

Answer

Applying one-hot encoding to numerical features.

Question 4

A data scientist is performing EDA on a dataset with 1,000 features and 10,000 rows. The target is binary. The scientist wants to reduce dimensionality while preserving information related to the target. Which TWO methods are appropriate?

Accepted Answer

L1-regularized logistic regression. L1-regularized logistic regression (C) is correct because the L1 penalty drives many of the 1,000 coefficients to exactly zero, performing embedded feature selection that keeps only the features contributing to predicting the binary target. Mutual information-based feature selection (D) is correct because it directly scores each feature by its statistical dependence on the binary target, allowing the scientist to retain the most target-relevant features while discarding the rest. PCA (A) is unsupervised and maximizes variance rather than target-related information, so it can discard directions that matter for classification. Autoencoders (B) are also unsupervised compression methods that optimize reconstruction, not target preservation. t-SNE (E) is a nonlinear visualization technique for embedding into 2-3 dimensions and is not a feature-selection or dimensionality-reduction method suitable for preserving target information.

Answer

Principal Component Analysis (PCA)

Answer

Autoencoders

Answer

t-Distributed Stochastic Neighbor Embedding (t-SNE)

Question 5

Refer to the exhibit. The training job 'my-job' failed with the error 'Unable to pull image from ECR'. What is the most likely cause?

Accepted Answer

The IAM role does not have permission to pull images from the ECR repository.. The error 'Unable to pull image from ECR' indicates that the SageMaker training job could not retrieve the custom Docker image stored in Amazon ECR. The most likely cause is that the IAM role associated with the training job lacks the `ecr:GetDownloadUrlForLayer` and `ecr:BatchGetImage` permissions required to pull images from the ECR repository. Without these permissions, SageMaker cannot authenticate and download the container image, even if the repository and image exist.

Answer

The instance type ml.m5.large does not support custom images.

Answer

The S3 bucket for training data is in a different account.

Answer

The role ARN is incorrect.

Question 6

A data scientist is exploring a dataset containing customer transaction records. The target variable is 'churn' (1 = churned, 0 = not churned). Which TWO actions should the scientist take to understand the data distribution and prepare for modeling?

Accepted Answer

Plot the frequency of the target variable to check for class imbalance.. Option C is correct because plotting the frequency of the target variable 'churn' is a standard exploratory data analysis step that reveals class imbalance, which directly affects modeling choices such as resampling, class weights, or appropriate evaluation metrics like F1-score or AUC-PR instead of accuracy. Option D is correct because checking for missing values in each column and deciding on an imputation strategy is essential data preparation; missing data can bias results or break algorithms, and the imputation method (mean, median, mode, or model-based) must be chosen before training. Option A is not appropriate at this stage because PCA is a dimensionality reduction technique typically applied after preprocessing and encoding, and it can obscure interpretability during initial exploration. Option B is premature because training a gradient boosting model to find important features is a modeling step, not an initial distribution-understanding or preparation action. Option E, while a valid preprocessing step for categorical variables, is not one of the two best actions for understanding data distribution and general preparation, and it should follow missing-value handling and target inspection.

Answer

Apply Principal Component Analysis (PCA) to reduce dimensionality.

Answer

Train a gradient boosting model to identify important features.

Answer

Convert categorical variables into one-hot encoded vectors.

Question 7

Refer to the exhibit. A data scientist wants to update the endpoint to use a new model image. The scientist updates the endpoint configuration with the new image and calls UpdateEndpoint. After the update, the endpoint status is 'Updating' but remains in that state for a long time. What is the most likely cause?

Accepted Answer

The new model image is failing health checks. When an endpoint update stalls in 'Updating' status, the most likely cause is that the new model image is failing health checks, so SageMaker cannot bring the new instances into service. The endpoint remains in 'Updating' while it retries or waits for healthy instances. This is a common failure mode when the model container doesn't respond correctly to the /ping health check endpoint.

Answer

The endpoint is already InService, so it cannot be updated

Answer

The old model image is no longer available

Answer

The instance count is too low to deploy both variants

Question 8

A data scientist is training a model using Amazon SageMaker with a custom Docker container. The training job fails with an error: 'Resource exhausted: Out of memory'. The training data is stored in S3. What should the data scientist do to resolve this issue?

Accepted Answer

Increase the instance memory by selecting a larger instance type.. The 'Resource exhausted: Out of memory' error indicates that the training instance's RAM is insufficient for the workload. Selecting a larger instance type with more memory directly addresses the OOM condition by providing additional physical RAM for model parameters, data batches, and intermediate computations. In SageMaker, instance types like ml.p3.2xlarge (61 GB RAM) vs. ml.p3.8xlarge (244 GB RAM) offer different memory capacities, and upgrading resolves memory exhaustion without altering the training logic.

Answer

Increase the EBS volume size attached to the training instance.

Answer

Use Pipe mode for data loading instead of File mode.

Answer

Reduce the batch size in the training script.

Question 9

A data scientist is training a time series forecasting model using Amazon SageMaker's DeepAR algorithm. The dataset contains daily sales data for 10,000 products over 2 years. The scientist splits the data chronologically: training on the first 18 months, validation on the next 3 months, and test on the last 3 months. The model performs well on validation but poorly on test. The data scientist suspects the model is overfitting to the validation period. Which action should the scientist take to improve test performance?

Accepted Answer

Use time series cross-validation with an expanding window. Time series cross-validation with an expanding window evaluates the model on multiple validation periods, ensuring robustness and reducing overfitting to a single validation window. Option B reduces context length and may lose long-term dependencies. Option C may introduce irrelevant features and does not address overfitting to validation. Option D removes validation entirely, which would not help diagnose or reduce overfitting.

Answer

Reduce the context length to 30 days

Answer

Add more exogenous features like holidays and promotions

Answer

Use the entire dataset for training and ignore validation

Question 10

A data engineer needs to transfer 50 TB of historical data from an on-premises Hadoop cluster to Amazon S3. The on-premises network has a 100 Mbps connection to AWS. The transfer must be completed within one week. Which approach should the engineer use?

Accepted Answer

Use AWS Snowball Edge device to physically transfer the data.. The on-premises network has a 100 Mbps connection, which yields a theoretical maximum transfer of about 1.08 TB per day (100 Mbps * 86400 seconds / 8 bits per byte / 1024^4 bytes per TB). To transfer 50 TB within 7 days, the required throughput is approximately 7.14 TB per day, far exceeding the available bandwidth. AWS Snowball Edge provides a physical shipping method that bypasses network constraints entirely, making it the only viable option for this volume and timeline.

Answer

Use Amazon S3 Transfer Acceleration.

Answer

Use AWS DataSync to transfer the data over the network.

Answer

Use multiple concurrent AWS CLI copy commands over VPN.

Question 11

A data scientist is performing feature selection for a linear regression model. Which TWO methods are appropriate? (Choose TWO.)

Accepted Answer

Lasso (L1) regularization. Lasso (L1) regularization (A) is appropriate because its L1 penalty drives some coefficients exactly to zero, effectively performing feature selection by shrinking irrelevant predictors out of the linear regression model. Forward selection (D) is also appropriate because it is a classic wrapper-based feature selection method that iteratively adds the predictor that most improves the model until a stopping criterion is met. Ridge (L2) regularization (B) is not a selection method since it only shrinks coefficients toward zero without eliminating any features. t-SNE (C) is a nonlinear dimensionality reduction and visualization technique, not a feature selection method for linear regression. PCA (E) performs feature extraction by creating new uncorrelated components rather than selecting the original features, so it does not meet the feature selection requirement.

Answer

Ridge (L2) regularization

Answer

t-distributed stochastic neighbor embedding (t-SNE)

Answer

Principal component analysis (PCA)

Question 12

A team is using Amazon SageMaker to train a linear regression model on a dataset with 10 features. After training, they notice the model has high bias. Which action is MOST likely to reduce bias?

Accepted Answer

Add polynomial features to capture non-linear relationships. High bias means the model is underfitting — it is too simple to capture the underlying relationship in the data. Adding polynomial features increases the model's capacity to represent non-linear relationships, which directly reduces bias. Linear regression on 10 raw features may be too rigid if the true relationship is curved.

Answer

Increase the regularization parameter lambda

Answer

Add L2 regularization

Answer

Use a smaller training dataset

Question 13

A data scientist is using Amazon SageMaker Autopilot to automatically build a model for a regression problem. The dataset has 100 features and 50,000 rows. Autopilot recommends a model with an R² of 0.85 on the validation set. However, when deployed to production, the model performs poorly (R² of 0.2). What is the most likely cause?

Accepted Answer

The production data distribution has shifted from the training data distribution. A large gap between validation performance (R² 0.85) and production performance (R² 0.2) with no change to the model itself is the classic signature of data drift — the production input distribution no longer matches the training/validation distribution. Autopilot's validation set is a random holdout from the same historical dataset, so it cannot detect drift that occurs after deployment. The fix is to monitor for drift (e.g., SageMaker Model Monitor) and retrain on recent data.

Answer

The model is overfitting to the training data

Answer

The model is underfitting due to insufficient training

Answer

Autopilot selected the wrong features

Question 14

Refer to the exhibit. A SageMaker training job uses an IAM role with this policy. The training job writes output to s3://my-bucket/output/. Which statement about the policy is true?

Accepted Answer

The training job can write output objects only if server-side encryption with SSE-S3 is used. The policy includes a Deny statement that explicitly denies PutObject requests unless the request includes the `x-amz-server-side-encryption` header set to `AES256`, which corresponds to SSE-S3. The Allow statement grants PutObject permission, but the Deny statement overrides it for any request that does not meet the encryption condition. Therefore, the training job can only write output objects if server-side encryption with SSE-S3 is used.

Answer

The Allow statement allows all PutObject requests regardless of encryption

Answer

The Deny statement blocks all PutObject requests

Answer

The GetObject permission requires the object to be encrypted with SSE-S3

Question 15

A financial services company uses Amazon SageMaker to train a fraud detection model. The training data is stored in an S3 bucket encrypted with AWS KMS. The SageMaker training job is configured to use a custom Docker container that reads data from S3 and writes model artifacts back to S3. The training job fails with the error: 'Unable to write model artifact to s3://my-bucket/output/model.tar.gz. Access Denied.' The IAM role used by the training job has the following permissions: s3:GetObject and s3:PutObject on the bucket, and kms:Decrypt on the KMS key. The training job is not using a VPC. What is the MOST likely cause of the failure?

Accepted Answer

The IAM role does not have kms:GenerateDataKey permission on the KMS key. The training job needs kms:GenerateDataKey permission to write objects encrypted with the KMS key. The provided IAM role has s3:GetObject, s3:PutObject, and kms:Decrypt, but lacks kms:GenerateDataKey, causing the Access Denied error when writing. Option A is incorrect because a region mismatch would not necessarily cause an Access Denied error if cross-region access is allowed, and the error indicates a permissions issue. Option C is incorrect because S3 Batch Operations is not required for writing artifacts. Option D is incorrect because the role already includes s3:PutObject.

Answer

The S3 bucket is in a different region than the training job

Answer

The S3 bucket requires S3 Batch Operations for writing artifacts

Answer

The IAM role does not have s3:PutObject permission on the output bucket

Question 16

A company ingests application logs into Amazon Kinesis Data Firehose and delivers them to Amazon S3. Compliance requires that each delivered object be encrypted with a customer-managed AWS KMS key, and that the Firehose service be able to write to the bucket without granting broad permissions. Which two actions must the data engineer take? (Choose two.)

Accepted Answer

Configure the Firehose delivery stream with a KMS key for SSE-KMS server-side encryption of the delivered S3 objects.. Delivering KMS-encrypted objects requires selecting a customer-managed key on the Firehose stream so SSE-KMS is applied at write time, and because Firehose assumes a delivery role, the destination bucket policy must authorize that role to put objects into the target prefix. Together these two controls provide the encryption and the least-privilege write access the compliance rule demands.

Answer

Create a VPC endpoint for Amazon S3 and route all Firehose traffic through it.

Answer

Enable S3 Object Lock in governance mode on the destination bucket.

Answer

Set the Firehose destination to Amazon OpenSearch Service and enable fine-grained access control.

Question 17

A data engineering team loads JSON event files into an AWS Glue Data Catalog table backed by Amazon S3. New files arrive continuously and each file contains records with a schema that occasionally gains new optional fields. A nightly AWS Glue ETL job reads the table and writes Parquet output. The job sometimes fails with an error that the schema does not match the table definition. The team wants the job to tolerate the new optional fields without rewriting historical data. What should the team do?

Accepted Answer

Set the job's catalog table update behavior to update the Data Catalog schema from the source, and enable schema evolution on the DynamicFrame reader.. AWS Glue readers can merge schemas observed across input files when schema evolution is enabled, and the job can push the reconciled schema back to the Data Catalog. That combination lets newly added optional fields flow through the job while older Parquet output stays valid, because added columns read as null for historical rows. No source data needs to be reprocessed.

Answer

Run an AWS Glue crawler with the 'Ignore the change and don't update the Data Catalog' option and continue the nightly job unchanged.

Answer

Increase the number of Glue DPUs allocated to the nightly job and enable job bookmarks.

Answer

Convert the source JSON files to Parquet first, then register the Parquet output in the Data Catalog for the nightly job to read.

Question 18

Which TWO actions can reduce inference latency for a SageMaker real-time endpoint? (Choose 2.)

Accepted Answer

Choose a larger instance type with more compute capacity.. Choosing a larger instance type with more compute capacity (Option A) reduces inference latency because it provides more CPU/GPU and memory resources, enabling the model to process each request faster. This directly decreases the time per inference, which is the primary driver of latency for real-time endpoints.

Answer

Add more instances behind the endpoint.

Answer

Use batch transform instead.

Answer

Switch to asynchronous inference.

Question 19

A machine learning engineer is analyzing a dataset with a mix of categorical and numerical features. The engineer wants to understand the correlation between categorical features and the target variable. Which statistical test is most appropriate for measuring association between a categorical feature and a binary target?

Accepted Answer

Chi-squared test of independence. The chi-squared test of independence is the standard statistical test for measuring association between two categorical variables, including a categorical feature and a binary target. It compares observed frequencies in a contingency table against expected frequencies under the null hypothesis of independence. A significant result indicates the feature and target are associated.

Answer

Pearson correlation coefficient

Answer

ANOVA (Analysis of Variance)

Answer

Mutual information

Question 20

A data scientist is training a deep learning model on Amazon SageMaker using a large dataset stored in S3. The training job is taking too long due to high I/O latency waiting for data to be downloaded from S3. Which action would MOST effectively reduce the I/O latency?

Accepted Answer

Use Pipe mode for the training channel. Pipe mode streams data directly from S3 into the training algorithm without writing to disk, eliminating the I/O latency caused by downloading files to the local storage. This is the most effective solution because the bottleneck is data transfer from S3, and Pipe mode reduces it to near-zero latency by feeding data on the fly.

Answer

Use File mode for the training channel

Answer

Increase the number of training instances

Answer

Use Amazon SageMaker Elastic Inference