MLS-C01 Mock Exam 74 — 25 Questions

Question 1

Match each SageMaker feature to its description.

Accepted Answer

Ground Truth: Build and manage training datasets with human labeling. The correct matches are: Ground Truth for dataset labeling, Neo for model optimization, Debugger for training monitoring, and Autopilot for automated model building. Common confusions include swapping Ground Truth and Neo, or mixing Debugger with Autopilot.

Answer

Ground Truth: Optimize models for deployment on target hardware

Answer

Neo: Build and manage training datasets with human labeling

Question 2

A data scientist is performing exploratory data analysis on a dataset with missing values. The dataset contains a column 'income' with 20% missing values. The income distribution is right-skewed. Which imputation method is most appropriate to preserve the skewness?

Accepted Answer

Impute with the median income. The median is robust to the right-skewed distribution of income. Imputing with the median preserves the skewness and central tendency without being influenced by outliers, unlike the mean which would pull the imputed values toward the tail and reduce skewness. Option A is wrong because the mean is sensitive to outliers and would distort the distribution. Option C is wrong because dropping rows reduces sample size and may bias the dataset. Option D is wrong because the mode is typically used for categorical data and is not meaningful for continuous skewed data.

Answer

Impute with the mean income

Answer

Drop rows with missing income

Answer

Impute with the mode income

Question 3

A company is using Amazon SageMaker to host a real-time inference endpoint for a natural language processing model. The endpoint is configured with an ml.m5.large instance. After deployment, the company observes that the inference latency is higher than expected, and the endpoint is experiencing CPU utilization near 100% during peak hours. The model is a PyTorch model that uses a transformer architecture. The company wants to reduce latency without increasing cost significantly. Which approach should the company take?

Accepted Answer

Change the endpoint instance type to ml.g4dn.xlarge to use GPU acceleration.. The issue is high CPU utilization causing latency for a PyTorch transformer model. GPU instances like ml.g4dn.xlarge can significantly accelerate inference through parallel processing, reducing latency. Option D is correct. Elastic Inference (C) may provide some acceleration but is less effective for transformer models and adds complexity. Auto Scaling (A) helps with traffic spikes but does not reduce per-request latency. Batch Transform (B) is for offline inference, not real-time.

Answer

Configure the endpoint with Auto Scaling to add more instances during peak hours.

Answer

Switch to batch transform for inference.

Answer

Attach an Elastic Inference accelerator to the existing instance.

Question 4

A machine learning engineer is deploying a model to an Amazon SageMaker endpoint. The model is a PyTorch model that requires a custom inference script. The engineer notices that the endpoint is returning 500 errors after deployment. Which step should the engineer take to debug the issue?

Accepted Answer

View the CloudWatch Logs for the endpoint.. When a SageMaker endpoint returns 500 errors, it typically indicates an application-level failure in the inference container, such as an unhandled exception in the custom inference script. CloudWatch Logs capture the stdout and stderr of the container, including Python tracebacks, which directly reveal the root cause. Option D is correct because viewing these logs is the first and most effective step to diagnose the specific error.

Answer

Redeploy the endpoint with a different instance type.

Answer

Check the CloudWatch metrics for the endpoint.

Answer

Modify the inference script and update the endpoint.

Question 5

An IAM policy attached to a SageMaker execution role is shown. A training job executed with this role fails with an error that the role cannot access the S3 bucket. The training job uses input data from s3://my-bucket/train/data.csv and output to s3://my-bucket/output/. What is the most likely cause?

Accepted Answer

The training job does not have s3:PutObject permission for the output location. The error message indicates the role cannot access the S3 bucket, which typically occurs when the role lacks write permissions to the output location. The training job needs s3:PutObject permission to write the output artifacts (model, logs, etc.) to s3://my-bucket/output/. Without this permission, SageMaker fails to save the training results, resulting in an access error.

Answer

The training job does not have s3:GetObject permission for the input data

Answer

The training data is encrypted with SSE-KMS and the role lacks KMS permissions

Answer

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

Question 6

A data engineering team needs to schedule a nightly ETL job that extracts data from an Amazon RDS for PostgreSQL instance, transforms it using Spark, and loads it into Amazon S3. The team wants to use AWS Glue for this task. Which components are required? (Select TWO.)

Accepted Answer

An AWS Glue ETL job with a Spark script.. An AWS Glue ETL job with a Spark script is required because the transformation step explicitly uses Spark. AWS Glue provides a managed Spark runtime, and the ETL job definition must include a script (either auto-generated or custom) that performs the extract, transform, and load operations. Without this component, the team cannot execute the Spark-based transformation logic.

Answer

An AWS Glue crawler to populate the Data Catalog.

Answer

An AWS Glue development endpoint.

Answer

An AWS Glue notebook for data exploration.

Question 7

A company is using Amazon SageMaker to deploy a model for real-time inference. The model has a latency requirement of less than 100 milliseconds. During testing, the latency is around 150 milliseconds. Which action can most likely reduce the latency to meet the requirement?

Accepted Answer

Use a larger instance type for the endpoint.. Enabling data capture adds overhead and increases latency. Using a larger instance type would provide more compute and reduce latency, but may increase cost. Reducing the batch size for inference (if batching is used) can reduce latency because the model processes fewer requests at once. However, the question implies a real-time endpoint which typically processes one request at a time; batch size might be 1. Increasing the variant weight for the production variant is for traffic routing, not latency. The most direct is to use a more powerful instance type. But also consider that increasing batch size (if using multi-record) increases latency. Reducing batch size reduces latency. However, for a real-time endpoint, the instance type is key. I'll go with using a larger instance type.

Answer

Reduce the batch size for inference.

Answer

Enable data capture for the endpoint.

Answer

Increase the initial variant weight for the production variant.

Question 8

A company is using Amazon SageMaker to train a model on a large dataset stored in S3. The training job is taking a long time due to slow data loading. Which action can the data scientist take to reduce data loading time?

Accepted Answer

Use Pipe mode to stream data from S3.. Pipe mode streams data directly from S3 into the training algorithm without first downloading it to the local storage, eliminating the bottleneck of disk I/O and reducing data loading time. This is especially effective for large datasets where the time to copy data to EBS (File mode) dominates the training job duration.

Answer

Use File mode and copy data to Amazon EBS.

Answer

Use a larger instance type with more memory.

Answer

Enable data augmentation during training.

Question 9

A company uses Amazon SageMaker to train a time-series forecasting model using the built-in DeepAR algorithm. The training data consists of daily sales for 1000 products over 2 years. The model performs well on most products, but for a few products with intermittent demand (sporadic sales), the predictions are poor. Which action should the data scientist take to improve predictions for these products?

Accepted Answer

Create a separate forecasting model specifically for intermittent demand products, using a model designed for such patterns (e.g., Croston's method).. Intermittent demand patterns (sporadic sales) require specialized models like Croston's method, which are designed to handle non-continuous demand. Option B is wrong because a simple linear regression model cannot capture the irregular spikes of intermittent demand; such models assume continuous, steady patterns. Option C is wrong because increasing the context length of DeepAR, which is built for regular time series, does not address the fundamental issue of sporadic demand—the model still expects continuous values. Option D is wrong because adding unrelated product categories introduces noise and does not help the model learn the specific intermittent pattern of the target products.

Answer

Use a linear regression model for all products.

Answer

Increase the context length of the DeepAR model to capture longer history.

Answer

Add more training data by including additional product categories.

Question 10

A data scientist is training a regression model on a dataset with 50 features. After training a linear regression model, the model achieves an R-squared of 0.85 on the training set but only 0.55 on the test set. Which technique is most likely to reduce the generalization error?

Accepted Answer

Apply L2 regularization (Ridge regression). The model exhibits high variance (overfitting): high training R² (0.85) but much lower test R² (0.55). L2 regularization (Ridge regression) shrinks coefficients toward zero, reducing model complexity and penalizing large weights, which directly combats overfitting and improves generalization to unseen data.

Answer

Add more features

Answer

Remove highly correlated features

Answer

Increase the polynomial degree of the model

Question 11

A data scientist is training a deep learning model for image classification using Amazon SageMaker. The training job is taking too long. The data scientist notices that GPU utilization is low (around 30%). Which action is most likely to improve GPU utilization and reduce training time?

Accepted Answer

Increase the batch size. Low GPU utilization (around 30%) indicates that the GPU is spending too much time idle while waiting for data batches to be processed. Increasing the batch size allows each training step to process more samples per forward/backward pass, which increases computational load on the GPU and improves hardware utilization. This directly reduces the number of steps needed per epoch, thereby decreasing overall training time.

Answer

Use a smaller instance type

Answer

Increase the learning rate

Answer

Reduce the batch size

Question 12

A company needs to move 50 TB of data from an on-premises data center to Amazon S3. The company has a limited internet bandwidth of 100 Mbps. The data transfer must be completed within 10 days. Which TWO services should the company use together to meet these requirements?

Accepted Answer

Amazon S3 as the destination. The correct answers are A (Amazon S3 as the destination) and D (AWS Snowball Edge). The company needs to transfer 50 TB within 10 days, but with only 100 Mbps bandwidth, only about 10.8 TB can be transferred over the internet in that time. Therefore, a physical transfer solution like AWS Snowball Edge is required to move the data offline to S3. Option B (AWS Direct Connect) is incorrect because even though it provides a dedicated connection, it still requires time to provision and would not be able to transfer 50 TB within 10 days if the bandwidth is limited to 100 Mbps (or even higher, Snowball is more practical for such large data). Option C (AWS Site-to-Site VPN) is incorrect because it also uses the internet and is subject to the same bandwidth limitations. Option E (AWS DataSync over the internet) is incorrect because it is also limited by the 100 Mbps bandwidth and cannot complete the transfer within 10 days.

Answer

AWS Direct Connect

Answer

AWS Site-to-Site VPN

Answer

AWS DataSync over the internet

Question 13

A data scientist wants to understand the distribution of a categorical feature with 100 unique values. Which visualization is most appropriate?

Accepted Answer

Bar chart. A bar chart is the most appropriate visualization for displaying the distribution of a categorical feature with 100 unique values because it uses discrete bars to represent the frequency or proportion of each category. Unlike a histogram, which requires continuous numeric bins, a bar chart preserves the distinct categories and allows clear comparison of counts across all 100 levels.

Answer

Histogram

Answer

Scatter plot

Answer

Pie chart

Question 14

A data engineer is using AWS Glue to catalog a dataset with 200 columns. During exploratory data analysis, they run a crawler and then view the table schema in the AWS Glue Data Catalog. They notice that many columns are inferred as 'string' even though they contain numeric values. What is the most likely cause?

Accepted Answer

The crawler sample size is too small, and the sampled rows contain non-numeric values.. The AWS Glue crawler samples a subset of rows to infer schema. If the sample size is too small or the sampled rows contain non-numeric values (e.g., headers, missing data, or text entries), the crawler may default to 'string' type for columns that are actually numeric. Options A, C, and D are incorrect: JSON files can contain numeric types, Parquet files support numeric types, and special characters in column names do not affect type inference.

Answer

The data is stored in JSON format, which only supports string types.

Answer

The data is stored in Parquet format, which does not support numeric types.

Answer

The column names contain special characters that prevent type inference.

Question 15

Which TWO of the following are benefits of feature scaling for machine learning algorithms?

Accepted Answer

Faster convergence of gradient descent. Feature scaling, typically via standardization (z-score) or min-max normalization, ensures that gradient descent converges faster. Without scaling, features with larger numerical ranges dominate the gradient updates, causing the algorithm to oscillate and require more iterations to reach the optimum. Scaling produces a more spherical contour of the loss function, allowing gradient descent to take more direct steps toward the minimum.

Answer

Eliminates the effect of outliers

Answer

Reduces the need for feature selection

Answer

Improves performance of decision tree algorithms

Question 16

A company is using Amazon Redshift for data warehousing. The data engineering team notices that queries are slow and the system is frequently writing to disk due to insufficient memory. Which type of workload management (WLM) configuration change would help reduce disk writes?

Accepted Answer

Increase the memory percentage allocated to the WLM queue.. When queries spill to disk in Amazon Redshift, it indicates that the memory allocated to the WLM queue is insufficient for the workload. Increasing the memory percentage for the queue allows more queries to be processed in memory, reducing the need to write intermediate results to disk and improving query performance.

Answer

Increase the number of query concurrency slots.

Answer

Enable query monitoring rules to abort queries that spill to disk.

Answer

Enable short query acceleration (SQA).

Question 17

Which TWO of the following are examples of unsupervised learning tasks?

Accepted Answer

Dimensionality reduction using PCA. Principal Component Analysis (PCA) is an unsupervised learning technique used for dimensionality reduction. It works by identifying the directions (principal components) that maximize variance in the data, without requiring any labeled target variable. This makes it a classic example of unsupervised learning, as the algorithm learns patterns solely from the input features.

Answer

Classifying emails as spam or not spam

Answer

Sentiment analysis of product reviews

Answer

Predicting house prices

Question 18

A company is using Amazon SageMaker to build a binary classification model. The dataset is highly imbalanced, with 95% negative class and 5% positive class. Which technique should be used to address the class imbalance?

Accepted Answer

Use a weighted loss function during training.. Using a weighted loss function (e.g., class weights in SageMaker's built-in XGBoost or custom PyTorch loss) assigns a higher penalty to misclassifications of the minority positive class. This directly addresses the 95:5 imbalance by making the model more sensitive to the positive class during gradient updates, without discarding data.

Answer

Use accuracy as the primary evaluation metric.

Answer

Perform random under-sampling of the majority class.

Answer

Remove all examples from the majority class.

Question 19

A data scientist needs to evaluate a binary classification model. The dataset is balanced. Which metric is most appropriate to compare model performance?

Accepted Answer

Accuracy. For a balanced binary classification dataset, accuracy is the most appropriate metric because it directly measures the proportion of correct predictions (true positives and true negatives) out of all predictions. Since the class distribution is equal, accuracy is not misleadingly high due to class imbalance, making it a reliable and straightforward measure of overall model performance.

Answer

Recall

Answer

F1 score

Answer

Precision

Question 20

During EDA, a data scientist discovers that two numerical features have a Pearson correlation coefficient of 0.95. Which action should the scientist take to avoid multicollinearity in a linear regression model?

Accepted Answer

Remove one of the features. Pearson correlation of 0.95 indicates high multicollinearity, which can adversely affect linear regression by inflating standard errors. Removing one of the correlated features (Option A) is a straightforward solution to eliminate multicollinearity. Option B (PCA) reduces dimensionality but creates principal components that are linear combinations, losing interpretability; it also does not directly remove the original features. Option C (Ridge regression) applies L2 regularization to shrink coefficients, which can mitigate multicollinearity but does not remove it; simply removing one feature is simpler. Option D (polynomial features) would introduce more correlated terms, worsening multicollinearity. Option E (min-max scaling) does not affect correlation. Therefore, removing one feature is the best action.

Answer

Apply PCA to the two features

Answer

Use Ridge regression to penalize coefficients

Answer

Create polynomial features from the correlated pair

Answer

Apply min-max scaling to both features