MLS-C01 Practice Exam 59 — 20 Questions

Question 1

An ML engineer runs the AWS CLI command above to list files in a training data bucket. The engineer notices that the three CSV files have different sizes but the same number of columns. What is the MOST likely cause of the size variation?

Choose an answer to begin — your selection is scored in the full session.

Accepted Answer

The files contain a different number of rows.. Different numbers of rows directly affect file size. Options A and B are incorrect: compression algorithms would cause size differences but the files are CSV and likely uncompressed; duplicate headers would cause schema inconsistency, but the question states same number of columns. Option D is incorrect because different column data types do not necessarily cause size variation; they could still have same number of rows.

Answer

The files are compressed with different algorithms.

Answer

Some files have duplicate headers.

Answer

The files have different column data types.

Question 2

A data scientist is exploring a dataset stored as a single 2 GB object in S3. The scientist wants to read only a subset of the file (e.g., the first 1000 lines) to perform initial data inspection. Which approach should the scientist take to minimize data transfer and cost?

Accepted Answer

Use S3 Select with a SQL query to retrieve the first 1000 rows.. S3 Select enables retrieving a subset of data using SQL queries, such as SELECT * FROM s3object LIMIT 1000, minimizing data transfer and cost. Option A is inefficient as it downloads the entire 2 GB file. Option C retrieves bytes, not lines, and may still transfer excessive data or require multiple requests to locate line boundaries. Option D scans the entire file with Athena, incurring cost and latency for a full table scan even with LIMIT 1000.

Answer

Use the AWS CLI to download the entire file and then use head to get the first lines.

Answer

Use the S3 Range header to read the first 1 MB of the file and parse lines.

Answer

Use Amazon Athena to query the file with LIMIT 1000.

Question 3

A machine learning team is working with a dataset containing high-dimensional sparse features, such as text data represented as bag-of-words. The team wants to reduce dimensionality while preserving the structure of the sparse matrix. Which technique is most appropriate for this scenario?

Accepted Answer

Truncated Singular Value Decomposition (SVD).. Truncated SVD (e.g., using sklearn's TruncatedSVD) is specifically designed for sparse matrices and efficiently reduces dimensionality while preserving the structure of the sparse matrix. Option A (t-SNE) is primarily for visualization and does not preserve the global structure well for dimensionality reduction. Option C (LDA) is a supervised method that requires labels and is not suitable for unsupervised dimensionality reduction. Option D (PCA using covariance matrix) requires a dense matrix and is computationally expensive for high-dimensional sparse data.

Answer

t-distributed Stochastic Neighbor Embedding (t-SNE).

Answer

Linear Discriminant Analysis (LDA).

Answer

Principal Component Analysis (PCA) using the covariance matrix.

Question 4

A data scientist is exploring a dataset with 100 features. The goal is to build a binary classification model. The dataset is highly imbalanced with 95% negative class and 5% positive class. The data scientist wants to understand the relationship between features and the target. Which technique is most appropriate for initial exploratory analysis?

Accepted Answer

Use stratified sampling to create a balanced subset for visualization and correlation analysis.. Stratified sampling preserves the class proportions, ensuring that the minority class (5% positive) is adequately represented in the subset for visualization and correlation analysis. Option A is wrong because removing the minority class would prevent any analysis of the target relationship. Option C is wrong because random sampling could miss the rare positive class entirely, leading to biased insights. Option D is wrong because SMOTE is a synthetic data generation technique intended for training, not for initial exploratory analysis.

Answer

Remove the minority class samples and analyze the majority class only.

Answer

Use random sampling to select 10% of the data for EDA.

Answer

Apply SMOTE to the dataset before performing EDA.

Question 5

A data scientist runs a SQL query on an Amazon Athena table and notices that the query scans a large amount of data. Which approach would reduce the amount of data scanned without changing the SQL logic?

Accepted Answer

Partition the table on a column that is frequently used in WHERE clauses.. Partitioning the table on a column that is frequently used in WHERE clauses allows Athena to prune partitions and only scan the relevant data, reducing the amount of data scanned. Option B (JSON) does not reduce scan because it is not columnar. Option C (Parquet without partitioning) is columnar and can reduce scan through column pruning, but without partitioning it still scans entire columns. Option D (GZIP) compresses data but Athena decompresses and scans the full file size, so no reduction in scanned data.

Answer

Convert the data from CSV to JSON format.

Answer

Store the data in Parquet format without partitioning.

Answer

Use GZIP compression on the data files.

Question 6

A data scientist is exploring a dataset with 100 features. After generating pair plots, the scientist notices that many features have skewed distributions. Which transformation should the scientist apply to make the distributions more Gaussian-like for modeling?

Accepted Answer

Yeo-Johnson transformation. (Yeo-Johnson transformation) is correct because it can handle both positive and negative values, making it suitable for datasets with skewed distributions that may include negative numbers. Option A (Log transformation) is wrong because it only works for strictly positive values. Option C (Standard scaling) does not change the shape of the distribution; it only centers and scales the data, so it does not fix skewness. Option D (Box-Cox transformation) also requires positive values, limiting its applicability.

Answer

Log transformation

Answer

Standard scaling (z-score normalization)

Answer

Box-Cox transformation

Question 7

A company is storing customer transaction data in Amazon S3 as CSV files. A data scientist uses AWS Glue to crawl the data and create a table in the AWS Glue Data Catalog. When querying the table with Amazon Athena, the data scientist notices that some columns have NULL values where data should exist. The data scientist examines the raw CSV files and confirms the data is present. What is the most likely cause of the NULL values?

Accepted Answer

The CSV files have different schemas (e.g., different columns) across partitions.. The most likely cause is that the CSV files have different schemas across partitions. When AWS Glue crawler infers the schema, it samples a subset of files. If partitions have different columns or column order, the inferred schema may not include columns present only in later partitions. When Athena queries the table, it uses the schema from the Data Catalog; columns missing from the schema appear as NULL. Option A is correct. Option B is incorrect because Athena does not skip corrupted records by default; it would fail on parse errors. Option C is incorrect because data type inference errors would cause different issues, such as type mismatches, not NULLs for existing data. Option D is incorrect because the Glue crawler can handle custom delimiters if configured; the issue here is schema mismatch, not delimiter recognition.

Answer

Athena is configured to skip corrupted records, causing NULLs.

Answer

The Glue crawler incorrectly inferred the data type of the columns.

Answer

The CSV files use a custom delimiter that the Glue crawler does not recognize.

Question 8

A company is using Amazon SageMaker to train a model. The training data is stored in an S3 bucket encrypted with AWS KMS. The SageMaker training role has the necessary permissions to decrypt the data. However, the training job fails with an access denied error. What is the most likely cause?

Accepted Answer

The training role does not have kms:Decrypt permission for the KMS key. The error message 'access denied' during a SageMaker training job with KMS-encrypted S3 data typically indicates that the training role lacks the kms:Decrypt permission for the specific KMS key used to encrypt the S3 objects. Even if the role has S3 read permissions (s3:GetObject), SageMaker must decrypt the data before reading it, which requires explicit KMS key policy or IAM policy granting kms:Decrypt. Without this, the training job fails with an access denied error.

Answer

The S3 bucket policy does not grant access to the training role

Answer

The training image is not compatible with encrypted data

Answer

CloudTrail logging is disabled

Answer

The training job is not in the same VPC as the S3 bucket

Question 9

Which TWO of the following are valid methods for handling missing values in a dataset before training a machine learning model?

Accepted Answer

Remove rows that contain missing values. Removing rows with missing values (listwise deletion) is a straightforward and valid method when the missing data is random and the dataset is large enough that the loss of rows does not significantly reduce statistical power or introduce bias. This approach ensures that only complete cases are used for training, avoiding the need to estimate missing values.

Answer

Use a decision tree algorithm that handles missing values internally

Answer

Increase the number of trees in a random forest

Answer

Replace missing values with zero

Question 10

Which THREE actions are valid steps in exploratory data analysis when working with a new dataset? (Choose three.)

Accepted Answer

Check the data types of each column.. Options A, B, and E are correct. A: Checking data types is fundamental in EDA to understand the nature of each variable. B: Generating descriptive statistics (mean, std, min, max) provides a quick summary of central tendency, dispersion, and range for numerical features. E: Creating histograms helps visualize the distribution of numerical features, revealing skewness, outliers, or patterns. Option C is incorrect because fitting a linear regression model is a modeling step, not part of EDA. Option D is incorrect because splitting the dataset into training and test sets is for model validation, not for initial data exploration.

Answer

Fit a linear regression model to identify important features.

Answer

Split the dataset into training and test sets.

Question 11

A data scientist is investigating an application that logs errors to Amazon CloudWatch Logs. The data scientist runs the CloudWatch Logs Insights query shown in the exhibit. The query returns no results, even though the data scientist knows errors have occurred. What is the most likely cause?

Accepted Answer

The filter pattern is case-sensitive and the log messages use a different case for 'error'.. CloudWatch Logs Insights queries are case-sensitive by default; the filter pattern 'ERROR' will not match log messages that use 'error' or 'Error'. Option A is incorrect because the stats count() function is spelled correctly in the query. Option C is incorrect because sorting by timestamp descending does not prevent results from being returned; it only affects the order. Option D is incorrect because bin(5m) is a valid function in CloudWatch Logs Insights when there are logs within the time range.

Answer

The stats count() function is misspelled.

Answer

The query sorts by timestamp descending, which hides results.

Answer

The bin(5m) function is not supported in CloudWatch Logs Insights.

Question 12

A data analyst is exploring a dataset with a binary target variable. Which TWO visualizations are most useful for understanding the relationship between a numerical feature and the target?

Accepted Answer

Histogram with overlaid target classes. Options C and D are correct. A histogram with overlaid target classes (C) allows viewing the distribution of the feature for each class, highlighting separability. A box plot grouped by target class (D) shows median, spread, and outliers per class, useful for comparing distributions. Option A (pie chart) is inappropriate for numerical features and binary targets. Option B (bar chart) is for categorical features, not numerical. Option E (scatter plot) requires two numerical variables; here the target is binary, so it would produce overlapping points and is less informative.

Answer

Pie chart of the feature

Answer

Bar chart of the feature

Answer

Scatter plot of the feature versus target

Question 13

A company is building a recommendation system using matrix factorization. The dataset has 1 million users and 100,000 items. The data scientist trains a model using SageMaker's Factorization Machines algorithm. The model achieves a root mean squared error (RMSE) of 0.95 on the test set. However, the business requires RMSE below 0.90. The data scientist has already tuned hyperparameters like number of factors and learning rate. Which additional step should the data scientist take to improve RMSE?

Accepted Answer

Add side features such as user demographics and item categories. (add side features) provides more information to the model, which can help reduce RMSE. Option B (more iterations) may not help if the model has already converged. Option C (increasing factors to 1000) could lead to overfitting. Option D (linear regression) is less powerful than matrix factorization for recommendation tasks.

Answer

Increase the number of training iterations

Answer

Increase the number of factors to 1000

Answer

Use a linear regression model instead

Question 14

A data scientist is analyzing a dataset with a mix of numerical and categorical features. The target variable is binary. The data scientist wants to visualize the distribution of a numerical feature across the two target classes. Which TWO visualization techniques are appropriate? (Choose 2.)

Accepted Answer

Overlapping histograms with transparency. Correct answers are options C and D. Option C (Overlapping histograms with transparency) is appropriate because it allows comparing the distribution of a numerical feature across two classes by overlaying histograms, making it easy to see shape, central tendency, and spread. Option D (Side-by-side boxplots) is appropriate because it succinctly displays median, quartiles, and outliers for each class, facilitating comparison. Option A (Heatmap of the correlation matrix) is not suitable as it visualizes correlations between features, not distribution of a single feature across classes. Option B (Stacked bar chart of the feature binned) is intended for categorical data, not numerical distributions. Option E (Scatter plot with color-coded classes) requires two numerical variables and is used to show relationships, not distribution of a single numerical feature.

Answer

Heatmap of the correlation matrix

Answer

Stacked bar chart of the feature binned

Answer

Scatter plot with color-coded classes

Question 15

A data scientist is exploring a dataset containing customer transactions. They want to create a feature that captures the average purchase amount per customer over the last 30 days. Which approach is most efficient in Amazon SageMaker Processing?

Accepted Answer

Use PySpark with window functions in SageMaker Processing. Using PySpark with window functions in SageMaker Processing allows efficient distributed computation for grouped time-series aggregations like average purchase amount per customer over the last 30 days. Option A is wrong because Amazon Athena SQL requires moving data out of SageMaker Processing and may not be as tightly integrated. Option C is wrong because iterating over rows with a Python for loop is inefficient and does not scale. Option D is wrong because pandas groupby and rolling functions may not scale to large datasets in a distributed environment; SageMaker Processing with PySpark provides better performance.

Answer

Use Amazon Athena SQL query with GROUP BY

Answer

Use a Python script with a for loop to calculate per customer

Answer

Use pandas groupby and rolling functions

Question 16

During exploratory data analysis, a data scientist notices that the distribution of a continuous feature is heavily right-skewed. Which transformation should be applied to make the distribution more symmetric for linear regression?

Accepted Answer

Log transformation. Log transformation is commonly used to reduce right skewness and make the distribution more symmetric. Standardization (z-score) does not change the shape of the distribution; it only centers and scales. One-hot encoding is for categorical features, not continuous. Min-max scaling also does not affect skewness; it rescales the range but preserves shape.

Answer

Standardization (z-score)

Answer

One-hot encoding

Answer

Min-max scaling

Question 17

During EDA, a data scientist generates a pairplot of the dataset and observes that two features have a Pearson correlation coefficient of 0.95. Which TWO conclusions can the scientist draw from this observation? (Choose 2)

Accepted Answer

The two features may be multicollinear. Options A and B are correct because a Pearson correlation coefficient of 0.95 indicates a very strong positive linear relationship between the two features. This strong linear relationship suggests potential multicollinearity if both features are used as predictors in a linear model. Option C is incorrect because a positive correlation means the features move in the same direction, not opposite. Option D is wrong because a high correlation implies dependence, not statistical independence. Option E is incorrect because correlation does not imply causation; it only measures the strength and direction of a linear relationship.

Answer

The two features move in opposite directions

Answer

The two features are statistically independent

Answer

One feature causes the other

Question 18

A machine learning engineer is deploying a custom XGBoost model for real-time inference on Amazon SageMaker. The model was trained using the SageMaker XGBoost built-in algorithm. The endpoint is deployed with an ml.m5.large instance and is receiving around 50 requests per second. The engineer notices that the endpoint's latency is around 200 ms, but the requirement is under 100 ms. The model's serialized format is a .tar.gz file. The engineer wants to reduce inference latency without modifying the model or retraining. What should the engineer do?

Accepted Answer

Use SageMaker Neo to compile the model for the target instance.. SageMaker Neo compiles trained models to optimize them for target hardware, reducing inference latency without modifying the model. Option A is wrong because SageMaker Debugger is used for monitoring training jobs and debugging, not for optimizing inference code. Option B is wrong because SageMaker Elastic Inference attaches GPU acceleration, which is beneficial for deep learning models but not for XGBoost (a tree-based model). Option D is wrong because SageMaker Batch Transform is designed for batch predictions on large datasets, not for real-time inference with low latency requirements.

Answer

Configure SageMaker Debugger to optimize the inference code.

Answer

Use SageMaker Elastic Inference to attach an accelerator.

Answer

Use SageMaker Batch Transform instead of a real-time endpoint.

Question 19

A data scientist is exploring a dataset with 50 features and a binary target. The data scientist computes the correlation matrix and finds that two features, X1 and X2, have a correlation coefficient of 0.95. Which TWO actions should the data scientist consider? (Choose 2.)

Accepted Answer

Remove one of the highly correlated features from the dataset.. Removing one of the highly correlated features reduces multicollinearity, which can stabilize model coefficients and improve interpretability. Option C is correct: Principal Component Analysis (PCA) transforms the correlated features into a set of uncorrelated components, effectively addressing multicollinearity. Option A is incorrect: Log transformation is used to handle skewness or scale differences, not correlation between features. Option D is incorrect: Creating an interaction term would add a new feature that is highly correlated with the original ones, potentially increasing multicollinearity. Option E is incorrect: Imputing missing values is unrelated to feature correlation; missing value imputation addresses data completeness, not multicollinearity.

Answer

Apply a log transformation to X1 and X2.

Answer

Create an interaction term between X1 and X2.

Answer

Impute missing values for X1 and X2.

Question 20

A data scientist is using Amazon SageMaker to train a model and wants to use a custom Docker container for training. The container requires access to a private Amazon ECR repository. Which IAM role configuration is needed?

Accepted Answer

Attach an IAM policy to the SageMaker execution role that allows ecr:GetDownloadUrlForLayer, ecr:BatchGetImage, and ecr:GetAuthorizationToken for the ECR repository.. The SageMaker execution role must have an IAM policy that includes ecr:GetDownloadUrlForLayer, ecr:BatchGetImage, and ecr:GetAuthorizationToken for the ECR repository. This is achieved by attaching an IAM policy to the SageMaker execution role, as described in Option A. Option B is incorrect because using the AWS account owner's role is not appropriate and would grant excessive permissions. Option C is incorrect because IAM users are not used for SageMaker execution roles; roles are used instead. Option D is incorrect because bucket policies apply to S3 buckets, not ECR repositories; ECR uses resource-based policies on the repository itself.

Answer

Use the AWS account owner's IAM role as the SageMaker execution role.

Answer

Create a new IAM user with ECR access and store credentials in SageMaker.

Answer

Add a bucket policy to the ECR repository allowing access from the SageMaker execution role.