MLA-C01 Data Preparation for Machine Learning Practice Test 3 — 15 Questions

Question 1

Refer to the exhibit. A SageMaker Processing job configured as above fails with a timeout error. The input data is 100 GB of CSV files. The processing script performs standard data cleaning operations. What is the most likely cause?

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

Accepted Answer

The processing job does not have enough memory for the data volume. The SageMaker Processing job is configured with a single `ml.m5.large` instance, which has 8 GiB of memory. The input data is 100 GB of CSV files, and the processing script performs standard data cleaning operations that typically load the entire dataset into memory (e.g., using pandas). With only 8 GiB of RAM, the instance cannot hold 100 GB of data, causing the job to run out of memory and eventually fail with a timeout error as the OS kills the process or the job hangs.

Answer

The container entrypoint is missing the full path to the script

Answer

The S3Input S3CompressionType is set to "None" but the file is compressed

Answer

The IAM role does not have permission to write to the output bucket

Question 2

Refer to the exhibit. A Glue job runs successfully the first time but on subsequent runs with new data (added to the same input location), the job does not process the new data. What is the most likely cause?

Accepted Answer

The job bookmark is enabled, causing the job to skip already processed data. When a Glue job bookmark is enabled, the job tracks previously processed data using a persistent state stored in a DynamoDB table. On subsequent runs, the bookmark mechanism skips files that have already been processed, so new data added to the same input location is ignored unless the bookmark is reset or the job is configured to process new partitions. This explains why the first run succeeds but later runs do not process new data.

Answer

The script location is incorrect

Answer

The MaxRetries is set to 0, so the job does not retry on failure

Answer

The WorkerType is Standard, which does not support incremental processing

Question 3

Refer to the exhibit. The Glue job reads a CSV file and attempts to write to a Parquet table. What is the most likely cause of this error?

Accepted Answer

The 'price' column contains non-numeric values in some rows. The error message indicates a 'NumberFormatException' when parsing the 'price' column, which occurs when Spark attempts to convert a string value to a numeric type. Since the Glue job's schema inference likely detected 'price' as a numeric column based on the majority of rows, any row containing a non-numeric value (e.g., 'N/A', 'null', or a currency symbol) will cause this parsing failure during the write to Parquet.

Answer

The 'price' column is missing from some rows

Answer

The schema inference incorrectly detected the column as String

Answer

The CSV file is compressed and not properly decompressed

Question 4

A company uses SageMaker Processing jobs to clean customer transaction data. The processing script runs on a single ml.m5.large instance and takes 30 minutes to process 50 GB of data in CSV format. To reduce processing time, the company wants to process 200 GB of data within 1 hour. Which combination of changes should the company make?

Accepted Answer

Increase InstanceCount to 4 and convert the data to Parquet format.. Increasing InstanceCount to 4 allows parallel processing of the 200 GB dataset across multiple ml.m5.large instances, each handling 50 GB, which directly reduces processing time. Converting the data from CSV to Parquet format further accelerates processing by enabling columnar storage and predicate pushdown, reducing I/O and CPU overhead. Together, these changes can achieve the goal of processing 200 GB within 1 hour, as the original 50 GB took 30 minutes on a single instance.

Answer

Run the job in local mode with a larger EBS volume.

Answer

Increase VolumeSizeInGB to 100 and use gzip compression.

Answer

Use a larger instance type (e.g., ml.r5.4xlarge) and keep the same script.

Question 5

A data scientist is preprocessing time series data for a fraud detection model. The data includes transaction timestamps, amounts, and merchant IDs. The model should predict fraud within seconds of a transaction. The data scientist wants to avoid data leakage by not using future information to predict past events. Which data preparation practice should be implemented?

Accepted Answer

Compute features like lagged transaction amounts and rolling statistics based only on each transaction's past data up to that point.. It ensures that features are computed using only historical data available up to each transaction's timestamp, preventing any future information from leaking into the model. In time series fraud detection, using only past data for lagged amounts and rolling statistics respects the temporal order and avoids the model learning patterns that would not be available at prediction time.

Answer

Randomly shuffle the dataset before splitting into training and validation sets.

Answer

Generate features such as rolling averages and lag features using a sliding window of all available data.

Answer

Normalize the features using MinMaxScaler on the entire dataset before splitting into training and testing.

Question 6

A team is collaborating on a machine learning project and needs to ensure that data used for training is consistent across experiments. The team wants to version datasets, track data lineage, and be able to reproduce past experiments. The team uses SageMaker for model training. Which combination of services and features should the team use?

Accepted Answer

Use SageMaker with AWS Lake Formation to manage data access, version datasets in S3, and use SageMaker Experiments to track training jobs.. It combines AWS Lake Formation for fine-grained data access control and governance, S3 versioning for dataset versioning, and SageMaker Experiments to track training jobs and lineage. This trio directly addresses the need for consistent data across experiments, versioning, lineage tracking, and reproducibility in SageMaker.

Answer

Use SageMaker Pipelines to automate training and store datasets in S3 with versioning enabled.

Answer

Store datasets in Amazon DynamoDB and use Amazon Athena to query specific versions.

Answer

Use S3 versioning to store all dataset versions and AWS Glue Data Catalog to track schema changes.

Question 7

A data scientist is using SageMaker built-in linear learner algorithm for a regression problem. The dataset has 10 features, some have missing values, and the target variable is right-skewed. The data scientist wants to handle missing values and transform the target variable to improve model performance. Which data preparation steps should the data scientist take?

Accepted Answer

Impute missing values with the median of each feature and apply a log transformation to the target variable.. Imputing missing values with the median is robust to outliers and preserves the distribution of each feature, which is important when the target is right-skewed. Applying a log transformation to the right-skewed target variable helps normalize its distribution, which aligns with the linear learner algorithm's assumption of normally distributed errors and improves convergence and prediction accuracy.

Answer

Apply one-hot encoding to all features and remove missing values by dropping rows.

Answer

Standardize all features to have zero mean and unit variance, then apply a box-cox transformation to the target.

Answer

Remove rows with missing values and normalize the target to range [0,1].

Question 8

A data team is preparing data for a machine learning pipeline. Which TWO practices are best for ensuring data quality and reproducibility? (Choose two.)

Accepted Answer

Implement automated data validation checks to catch anomalies in new data.. Option C is correct because automated data validation checks (e.g., schema enforcement, range/null checks, drift detection) systematically catch anomalies in incoming data, which directly ensures consistent data quality at scale in an ML pipeline. Option E is correct because storing cleaned and transformed datasets in S3 with versioning enabled preserves immutable, uniquely identified snapshots (via version IDs), which is essential for reproducibility—any model can be retrained against the exact dataset version used previously. Option A is not among the marked correct answers, and while a fixed random seed aids repeatability of sampling, it does not by itself ensure data quality or end-to-end reproducibility of the dataset artifacts. Option B is not marked correct because shuffling before splitting affects distribution and leakage concerns but does not guarantee data quality or reproducibility. Option D is not marked correct because manual inspection and outlier removal is subjective, non-scalable, and not reproducible compared to automated, versioned processes.

Answer

Use a fixed random seed when sampling data to ensure repeatability.

Answer

Shuffle the dataset before splitting into train and test sets.

Answer

Manually inspect and clean data to remove outliers.

Question 9

A data scientist is working with a dataset containing customer demographics and purchase history. The dataset includes categorical variables with high cardinality (e.g., ZIP code, product ID). The data scientist wants to perform feature engineering to improve model performance. Which THREE feature engineering techniques should the data scientist consider? (Choose three.)

Accepted Answer

Principal Component Analysis (PCA) to reduce dimensionality of numerical features.. Target encoding (C) replaces each high-cardinality category with a statistic derived from the target variable, such as the mean target value for that category, which compresses thousands of ZIP codes or product IDs into a single informative numeric column without exploding dimensionality. Frequency encoding (D) represents each category by how often it appears in the dataset, giving the model a compact numeric signal about rare versus common ZIP codes or product IDs while avoiding the curse of dimensionality. PCA (A) is also appropriate because it reduces the dimensionality of the numerical feature space by projecting correlated numeric variables onto orthogonal principal components, which can improve model performance and mitigate overfitting when many numeric demographic or purchase-history features are present. Domain-specific feature engineering (B) is not marked correct here because, although it is generally useful, the scenario specifically asks for techniques suited to high-cardinality categorical variables and dimensionality reduction, and it is not one of the three expected answers. One-hot encoding all categorical features (E) is not correct because one-hot encoding high-cardinality variables such as ZIP code or product ID creates an extremely sparse, very high-dimensional matrix that can degrade model performance and increase memory and training time.

Answer

Domain-specific feature engineering based on business rules.

Answer

One-hot encoding all categorical features.

Question 10

A data engineer needs to provide the data science team with access to various data sources for machine learning. The team uses Amazon SageMaker Studio. Which TWO data sources can be accessed directly from SageMaker Studio notebooks without additional infrastructure? (Choose two.)

Accepted Answer

Amazon S3.. Option A (Amazon S3) is correct because SageMaker Studio notebooks have native, built-in integration with Amazon S3 through the AWS SDK for Python (Boto3), the SageMaker Python SDK, and direct S3 URI paths, so no extra infrastructure is needed to read or write training data. Option E (Amazon Athena) is correct because Athena is a serverless query service that can be invoked directly from a Studio notebook via Boto3, the AWS Data Wrangler/PyAthena libraries, or SQL magic, querying data in S3 without provisioning any servers. Option B (Amazon Redshift) is not directly accessible without additional setup such as configuring a cluster, VPC connectivity, and the Redshift Data API or JDBC/ODBC drivers. Option C (Amazon DynamoDB) requires SDK calls and appropriate IAM/endpoint configuration rather than being a natively integrated, zero-infrastructure source in Studio. Option D (Amazon RDS for MySQL) requires network/VPC configuration and a database driver or connector, so it is not directly accessible without additional infrastructure.

Answer

Amazon Redshift.

Answer

Amazon DynamoDB.

Answer

Amazon RDS (MySQL).

Question 11

A data scientist needs to convert categorical variables to numerical format for a linear regression model. The dataset contains a 'Country' column with 50 unique values. Which transformation should the engineer use to avoid introducing ordinal relationships?

Accepted Answer

One-hot encoding. One-hot encoding is correct because it creates binary columns for each category, avoiding any implicit ordinal relationship between the 50 unique countries. This is essential for linear regression, which assumes numerical inputs have meaningful order; one-hot encoding ensures the model treats each country as an independent category without ranking.

Answer

Label encoding

Answer

Target encoding

Answer

Ordinal encoding

Question 12

A data engineer is processing a large dataset in Amazon S3 with AWS Glue ETL. The dataset contains timestamps in multiple time zones. The engineer needs to create a feature for hour-of-day consistent across all records. Which approach ensures correctness?

Accepted Answer

Convert all timestamps to UTC during the ETL process, then extract hour. Converting all timestamps to UTC during the ETL process ensures a consistent time zone reference before extracting the hour-of-day feature. This avoids ambiguity from mixed time zones and aligns with best practices for machine learning feature engineering. AWS Glue ETL with Apache Spark provides built-in functions like `to_utc_timestamp()` to perform this conversion reliably.

Answer

Convert all timestamps to UTC in the ETL script using Spark's from_utc_timestamp

Answer

Use AWS Glue's built-in transform to parse timestamps with timezone offsets

Answer

Use Python's datetime.strptime with tzlocal

Question 13

A machine learning engineer needs to handle missing values in a dataset containing numerical features. The missingness is completely at random (MCAR). Which imputation strategy is most robust for downstream model performance?

Accepted Answer

Impute with median of each feature. When missingness is completely at random (MCAR), imputing with the median is robust because it preserves the central tendency of the distribution without introducing bias or distorting variance. Unlike mean imputation, the median is resistant to outliers, making it a safe default for numerical features in downstream models that assume normally distributed inputs or are sensitive to skewed data.

Answer

Impute with a constant like -1

Answer

Use a model to predict missing values

Answer

Remove all rows with missing values

Question 14

A data scientist is preparing text data for natural language processing (NLP). The corpus contains many rare words and typos. To reduce dimensionality and improve generalization, they decide to apply stemming and remove stop words. However, after training, the model performs poorly on domain-specific terms. What is the most likely cause?

Accepted Answer

Both stemming and stop word removal are inappropriate for the domain. Both stemming and stop word removal are inappropriate for this domain. Stemming aggressively reduces words to their root forms, which can conflate distinct domain-specific terms (e.g., 'therapy' and 'therapist' both stem to 'therap'), losing critical semantic nuance. Stop word removal can discard words that carry domain-specific meaning (e.g., 'not' in medical negation or 'up' in 'tune-up' for maintenance), leading to poor generalization on specialized vocabulary.

Answer

The corpus should be lemmatized instead

Answer

Stemming is too aggressive for the domain

Answer

Stop word removal removed important context words

Question 15

A data scientist is using Amazon SageMaker Data Wrangler for feature engineering on a large dataset stored in S3. The dataset has a column 'ProductCategory' with 1000+ unique values. To reduce dimensionality, they want to group categories that appear less than 1% of the time into an 'Other' category. Which Data Wrangler transform should they use?

Accepted Answer

Handle rare values. The 'Handle rare values' transform in SageMaker Data Wrangler is specifically designed to group infrequent category values into a single 'Other' bucket based on a frequency threshold (e.g., less than 1%). This directly addresses the need to reduce dimensionality by consolidating rare categories without requiring custom code or manual grouping.

Answer

Group similar categories

Answer

Custom transform with Python

Answer

One-hot encode with threshold