MLS-C01 Practice Test 15 — 15 Questions

Question 1

A data scientist is training a recurrent neural network (RNN) for time series forecasting. The model's training loss is not decreasing, and the gradients are vanishing. Which technique should the scientist apply to address vanishing gradients?

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

Accepted Answer

Replace the RNN cells with LSTM or GRU units.. Replacing RNN cells with LSTM or GRU units directly addresses vanishing gradients because these architectures include gating mechanisms that regulate the flow of information, allowing gradients to propagate over many time steps without diminishing. LSTMs and GRUs are specifically designed to capture long-term dependencies, making them the standard solution for vanishing gradients in recurrent networks.

Answer

Apply gradient clipping.

Answer

Add batch normalization layers.

Answer

Increase the learning rate.

Question 2

A data scientist is working with a dataset that contains a feature with many outliers. Which transformation should the scientist apply to reduce the impact of outliers?

Accepted Answer

Log transformation. Log transformation compresses the range of values and reduces the impact of outliers. Standardization (z-score) does not reduce outlier impact. Min-max scaling is sensitive to outliers. Square root transformation is less effective than log for large outliers. Binning loses information.

Answer

Min-max scaling

Answer

Standardization (z-score)

Answer

Binning

Question 3

A data scientist is performing exploratory data analysis on text data. They want to identify the most common terms and their frequencies. Which approach should they use?

Accepted Answer

Create a bag-of-words matrix and compute term frequencies.. A bag-of-words matrix counts the frequency of each term in the text, directly providing the most common terms and their frequencies. Option A is incorrect because sentiment analysis determines the emotional tone of text, not term frequencies. Option B is incorrect because Latent Dirichlet Allocation (LDA) is a topic modeling technique that assigns topics to documents, not term frequencies. Option D is incorrect because word2vec generates dense vector embeddings that capture semantic relationships, not raw term frequencies.

Answer

Perform sentiment analysis on the text.

Answer

Apply Latent Dirichlet Allocation (LDA) to extract topics.

Answer

Use word2vec to generate word embeddings.

Question 4

A data scientist is building a binary classifier and wants to evaluate model performance. Which THREE metrics are most commonly used?

Accepted Answer

Precision. For a binary classifier, Precision (C) is correct because it measures the proportion of positive predictions that are truly positive (TP / (TP + FP)), which is essential when false positives matter. Recall (D) is correct because it measures the proportion of actual positives correctly identified (TP / (TP + FN)), which is essential when false negatives matter. Accuracy (E) is correct because it measures the overall fraction of correct predictions ((TP + TN) / total), a standard baseline metric for classification. Mean Absolute Error (A) and RMSE (B) are not correct here because they are regression metrics that quantify the average magnitude of continuous prediction errors, not classification performance.

Answer

Mean Absolute Error

Answer

RMSE

Question 5

A company is training a large language model on Amazon SageMaker using a single GPU instance. The training is taking too long. Which change would most likely reduce training time?

Accepted Answer

Use a larger instance with multiple GPUs and enable distributed training. Using multiple GPUs in a distributed training job can parallelize work and reduce time. Option B is wrong because increasing instance memory may not help if the bottleneck is GPU compute. Option C is wrong because decreasing batch size often increases training time due to more updates per epoch. Option D is wrong because S3 Glacier is designed for archival storage with slower retrieval, not faster access.

Answer

Increase the instance memory

Answer

Decrease the batch size

Answer

Store the training data in S3 Glacier for faster access

Question 6

A data scientist is exploring a dataset and finds that the correlation between two features is 0.95. What should the data scientist do to address multicollinearity before training a linear regression model?

Accepted Answer

Remove one of the two features. A correlation of 0.95 between two features indicates severe multicollinearity, which inflates the variance of linear regression coefficients and makes them unstable and hard to interpret. The most direct and standard remedy is to remove one of the two highly correlated features, eliminating the redundancy. This preserves model interpretability and avoids the coefficient instability that multicollinearity causes.

Answer

Apply L2 regularization

Answer

Standardize the features

Answer

Apply Principal Component Analysis

Question 7

A company uses Amazon Redshift for its data warehouse. The data engineering team notices that queries are slow and wants to improve performance without changing the schema. Which action is most likely to improve query performance?

Accepted Answer

Increase the number of nodes in the cluster.. Increasing the number of nodes in an Amazon Redshift cluster distributes data and query processing across more compute resources, which directly improves parallel execution and reduces query execution time. This is the most effective way to boost performance without altering the schema, as it scales the cluster's CPU, memory, and I/O capacity.

Answer

Decrease the number of nodes to reduce network overhead.

Answer

Disable compression on all tables to reduce CPU overhead.

Answer

Change the distribution style from AUTO to EVEN.

Question 8

A machine learning engineer is deploying a model that predicts customer churn. The model outputs probabilities between 0 and 1. The business requires that at least 90% of customers flagged for churn actually churn (precision >= 0.9). Currently, the model's precision is 0.85 at the default threshold of 0.5. Which threshold adjustment should the engineer consider?

Accepted Answer

Increase the threshold to 0.7. Increasing the threshold to 0.7 raises the probability cutoff for classifying a customer as churning. This means only customers with a high predicted probability (strong model confidence) are flagged, which reduces false positives and increases precision. Since the current precision at 0.5 is 0.85 and the goal is ≥0.9, moving the threshold higher is the correct direction to achieve the required precision.

Answer

Decrease the threshold to 0.4

Answer

Decrease the threshold to 0.3

Answer

Keep the threshold at 0.5

Question 9

A data scientist is training a binary classifier using a large dataset with class imbalance (90% negative, 10% positive). After training a logistic regression model, the F1 score is low but accuracy is high. Which TWO actions should the data scientist take to improve model performance? (Choose 2.)

Accepted Answer

Switch to evaluation metrics such as F1 score or AUC-ROC instead of accuracy.. Option A is correct because with a 90/10 class imbalance, accuracy is misleading — a trivial model predicting all negatives achieves 90% accuracy — so the data scientist should rely on imbalance-aware metrics like F1 score (harmonic mean of precision and recall) or AUC-ROC, which measures ranking quality across all thresholds and is insensitive to class prior. Option D is correct because resampling the training data with techniques such as SMOTE (Synthetic Minority Over-sampling Technique) generates synthetic minority-class examples via k-nearest-neighbor interpolation, balancing class priors and helping the model learn the positive class instead of defaulting to the majority. Option B is not the issue here: logistic regression is not distance-based, and feature scaling mainly affects convergence speed or regularized coefficient penalties, not the class-imbalance-driven low F1. Option C is wrong because adding features increases capacity and risks overfitting; it does not address the imbalance that causes the model to under-predict the minority class. Option E is wrong because increasing regularization shrinks coefficients and reduces variance, but the problem is not overfitting — it is the skewed decision boundary from imbalanced classes, which regularization would not fix.

Answer

Apply feature scaling to ensure all features contribute equally.

Answer

Add more features to the model to improve its capacity.

Answer

Increase the regularization parameter to reduce overfitting.

Question 10

A data scientist is working with a dataset that contains text reviews and a numeric rating (1-5). The goal is to predict the rating from the review text. During EDA, the scientist wants to check if there are any spelling errors or unusual characters. Which tool is BEST suited for this task?

Accepted Answer

Amazon Comprehend to detect syntax and entities.. Amazon Comprehend is the best choice among the options because it is an AWS AI service that can detect syntax, entities, and key phrases in text. While it does not directly find spelling errors, it can identify unusual patterns or anomalies in text that may indicate misspellings or odd characters. SageMaker Data Wrangler is for tabular data, Athena is for SQL queries, and QuickSight is for visualization, none of which are specialized for text analysis in this context.

Answer

Amazon SageMaker Data Wrangler with a custom transform for text cleaning.

Answer

Amazon Athena with SQL queries to find anomalies.

Answer

Amazon QuickSight to create word clouds.

Question 11

A company is using Amazon Kinesis Data Streams to ingest real-time clickstream data. The data is consumed by a Kinesis Data Analytics application that runs SQL queries. The application has been failing intermittently with 'ProvisionedThroughputExceededException' errors. Which action should be taken to resolve this issue?

Accepted Answer

Increase the number of shards in the Kinesis data stream.. The 'ProvisionedThroughputExceededException' error indicates that the Kinesis Data Stream's read or write throughput limits have been exceeded. Increasing the number of shards in the stream directly increases the total provisioned throughput, allowing the Kinesis Data Analytics application to consume data without throttling.

Answer

Disable error logging in the Kinesis Data Analytics application.

Answer

Increase the record size in the Kinesis data stream.

Answer

Switch from Kinesis Data Analytics to Kinesis Data Firehose.

Question 12

A data analyst is using AWS Glue to catalog datasets for exploratory analysis. The analyst wants to understand the schema and data types. Which TWO tools can the analyst use to view the schema of a table in the AWS Glue Data Catalog? (Choose TWO.)

Accepted Answer

Amazon Athena. Amazon Athena can query the AWS Glue Data Catalog using SQL, including viewing table schemas via the INFORMATION_SCHEMA or by running DESCRIBE statements on tables. The AWS Glue console directly displays the schema of tables in the Data Catalog under the 'Tables' section. Amazon Redshift query editor is for querying Redshift data warehouses, not for directly viewing Glue Catalog schemas unless a federated query is set up. Amazon QuickSight is a business intelligence tool for visualizing data, not for schema exploration. Amazon S3 console only shows objects in S3 buckets, not the schema of Glue tables.

Answer

Amazon Redshift query editor

Answer

Amazon QuickSight

Answer

Amazon S3 console

Question 13

A company uses Amazon Kinesis Data Firehose to deliver streaming data to Amazon S3. The data contains personally identifiable information (PII) that must be redacted before storage. Which AWS service can be integrated with Kinesis Data Firehose to transform the data in real time?

Accepted Answer

AWS Lambda. AWS Lambda can be integrated as a data transformation function within a Kinesis Data Firehose delivery stream. When Firehose receives incoming records, it can invoke a Lambda function synchronously to process each batch of data, allowing you to redact PII fields in real time before the data is written to Amazon S3.

Answer

Amazon Athena

Answer

Amazon Kinesis Data Analytics

Answer

Amazon EMR

Question 14

A data scientist is using Amazon SageMaker to build a custom training algorithm. The algorithm requires a specific library that is not included in the default SageMaker containers. The scientist wants to create a custom container that includes this library. Which TWO steps are required? (Choose TWO.)

Accepted Answer

Build a Docker image with the required library. To use a custom container in Amazon SageMaker, the scientist must first build a Docker image that installs the required library on top of a suitable base image, so option C is correct. The image must then be pushed to Amazon Elastic Container Registry (Amazon ECR), because SageMaker pulls training and inference images from ECR repositories, making option E correct. Uploading the Docker image to Amazon S3 (A) is not how SageMaker accesses container images; S3 is used for data and model artifacts, not for running Docker images. Creating an AWS Lambda layer (B) applies to Lambda functions, not SageMaker containers. Registering the container in the SageMaker Model Registry (D) is an optional model-governance step and is not required to build or use a custom training container.

Answer

Upload the Docker image to an Amazon S3 bucket

Answer

Create an AWS Lambda layer with the library

Answer

Register the container in the SageMaker Model Registry

Question 15

A data scientist needs to understand the distribution of a continuous variable in a large dataset stored in Amazon S3. Which AWS service is most appropriate for quickly generating summary statistics and visualizations?

Accepted Answer

Amazon QuickSight. Amazon QuickSight is a business intelligence service that connects to data sources including Amazon S3 (via Athena or direct S3 manifest) and provides fast, interactive visualizations and summary statistics. It is designed for analysts to build dashboards and explore distributions without writing complex code. This makes it the most appropriate choice for generating visualizations of a continuous variable.

Answer

AWS Glue

Answer

Amazon Athena

Answer

Amazon SageMaker Studio