MLS-C01 Practice Exam 58 — 20 Questions

Question 1

A data scientist is exploring a dataset and wants to check for missing values. Which method is most appropriate to identify the percentage of missing values per column?

Accepted Answer

Use pandas .isnull().sum() in a SageMaker notebook. Using pandas .isnull().sum() in a SageMaker notebook is the most appropriate method because it directly provides the count (and thus the percentage when divided by total rows) of missing values per column, which is a standard exploratory data analysis technique. Option A is incorrect because Amazon S3 Select is used for filtering and retrieving subsets of data from S3 objects, not for computing missing values. Option B is incorrect because while Amazon Athena can run SQL queries like SELECT COUNT(*), it is less direct for per-column missing value analysis and requires a schema. Option C is incorrect because Amazon QuickSight is a visualization tool, not designed for programmatic missing value detection. Option D is incorrect because AWS Glue Crawler discovers schema and partitions, not missing values.

Answer

Use Amazon S3 Select to query missing values

Answer

Use Amazon Athena to run a SELECT COUNT(*) query

Answer

Use Amazon QuickSight to create a missing value dashboard

Answer

Use AWS Glue Crawler to detect missing values

Question 2

A data scientist is analyzing a dataset with 1 million rows and 50 features. The scientist wants to detect outliers in a numerical feature 'transaction_amount' which has a long right tail. The scientist suspects that outliers are due to data entry errors and should be removed. Which outlier detection method is MOST robust for this scenario?

Accepted Answer

Interquartile range (IQR) with multiplier 1.5. The IQR method (Option A) is the most robust for detecting outliers in the 'transaction_amount' feature because it is based on quartiles and does not assume any underlying distribution, making it resistant to skew and extreme values. Z-score (Option C) is inappropriate because it assumes a normal distribution, which the long right tail violates. Mahalanobis distance (Option B) assumes multivariate normality and is not suited for univariate outlier detection. DBSCAN (Option D) is computationally expensive on 1 million rows and is designed for density-based clustering, not for univariate outlier detection.

Answer

Mahalanobis distance

Answer

Z-score with threshold 3

Answer

DBSCAN clustering

Question 3

A data engineer is exploring a large dataset in Amazon Athena. The dataset is partitioned by date and stored in Parquet format. The engineer wants to check the number of distinct values in a column for a specific date range. Which THREE practices reduce query cost and improve performance?

Accepted Answer

Use the COUNT(DISTINCT column) function.. Options A, B, and E are correct. Using COUNT(DISTINCT column) (A) is a precise way to count distinct values, and while it scans the column, it avoids fetching unnecessary data. Filtering with a WHERE clause on the partition column (B) limits the data scanned to only the relevant partitions, significantly reducing cost and improving performance. Using a columnar format like Parquet (E) reduces I/O by reading only the required columns. Option C (ORDER BY) is incorrect because it requires sorting the entire result set, increasing processing time and cost. Option D (SELECT *) is incorrect as it retrieves all columns, negating the benefits of columnar storage and increasing data scanned.

Answer

Use ORDER BY to sort the results.

Answer

Use SELECT * to retrieve all columns.

Question 4

A team is building a model to predict customer churn. They have 50 features, including categorical variables with high cardinality (e.g., zip code with 10,000 unique values). Which feature engineering technique is most appropriate?

Accepted Answer

Target encoding. Target encoding replaces each category with the mean of the target variable, which handles high cardinality well. Option A (binning) reduces cardinality but loses information. Option B is correct because target encoding is specifically designed for high-cardinality categorical features. Option C (label encoding) implies ordinality and can introduce misleading relationships. Option D (one-hot encoding) would create 10,000 binary columns, causing high dimensionality.

Answer

Binning zip codes into regions

Answer

Label encoding

Answer

One-hot encoding

Question 5

A machine learning team is using Amazon SageMaker to train a model. The training job uses spot instances to reduce cost. However, the training job is frequently interrupted. Which TWO actions can help mitigate the impact of spot interruptions? (Choose TWO.)

Accepted Answer

Use managed spot training with SageMaker's 'ManagedSpotTraining' parameter set to True.. Managed spot training (C) automatically manages the lifecycle of spot instances, including saving checkpoints and relaunching training when capacity becomes available. Checkpointing (D) saves model state periodically to Amazon S3, allowing training to resume from the last checkpoint after an interruption, minimizing progress loss. Option A (increasing instances) does not prevent interruptions and raises cost. Option B (larger instance type) does not guarantee lower interruption rates and is cost-inefficient. Option E (on-demand instances) avoids interruptions but defeats the purpose of cost reduction.

Answer

Increase the number of training instances.

Answer

Use a larger instance type that is less likely to be interrupted.

Answer

Switch to on-demand instances.

Question 6

A machine learning team is deploying a real-time inference endpoint on Amazon SageMaker for a model that requires low latency (<100 ms). The model is a PyTorch model with custom pre- and post-processing logic. The team uses a SageMaker Model with a custom inference container. After deployment, they observe that the endpoint takes over 500 ms for the first request, but subsequent requests are fast (~50 ms). What is the MOST likely cause?

Accepted Answer

The container has a cold start delay because the model needs to be loaded into memory from Amazon S3 on the first request.. The first request triggers a cold start where the custom inference container initializes and loads the model from Amazon S3 into memory, causing high latency. Subsequent requests are fast because the model remains cached. Option A is wrong because the instance type primarily affects throughput and steady-state latency, not transient cold starts. Option B is wrong because the issue is not about memory exhaustion—the endpoint handles subsequent requests well. Option D is wrong because auto-scaling adds instances but does not eliminate the cold start for the initial request on a new instance.

Answer

The instance type is too small to handle the model size.

Answer

The model is too large and exceeds the instance memory.

Answer

The endpoint is not configured with auto-scaling.

Question 7

A machine learning team is using SageMaker to train a model with a custom Docker container. The training script runs locally but fails on SageMaker with a 'Permission denied' error when writing to /opt/ml/model. What is the likely cause?

Accepted Answer

The container's user does not have write permission to /opt/ml/model. In SageMaker, the training container is expected to store the trained model artifacts in the /opt/ml/model directory. If the user running the training script inside the container does not have write permissions to that directory, the training will fail with a 'Permission denied' error. Option A is correct. Option B (image too large) would cause different errors, such as EBS volume limits. Option C refers to input data paths; the error is about writing the model, not reading inputs. Option D (S3 bucket) would cause read errors, not a write permission issue.

Answer

The Docker image is too large

Answer

The training script is trying to read from /opt/ml/input/data instead of /opt/ml/input/data/training

Answer

The training data is not in the correct S3 bucket

Question 8

An IAM policy attached to a SageMaker execution role is shown in the exhibit. When a data scientist tries to create a training job that writes logs to CloudWatch Logs, the job fails. What is the MOST likely reason?

Accepted Answer

The policy lacks permissions for CloudWatch Logs actions. The IAM policy attached to the SageMaker execution role does not include permissions for CloudWatch Logs actions (e.g., CreateLogGroup, CreateLogStream, PutLogEvents), which are required to write logs during training job execution. Option A is incorrect because the API version is not relevant to permission issues. Option B is incorrect because the problem is not related to S3 bucket policy; the training job fails due to logging, not data access. Option D is incorrect because the policy does not contain an explicit deny for SageMaker actions, but the lack of CloudWatch Logs permissions causes the failure.

Answer

The policy does not specify the SageMaker API version

Answer

The S3 bucket policy denies access to the training job

Answer

The policy has an implicit deny for SageMaker actions

Question 9

Which THREE of the following are common issues that can be identified during exploratory data analysis? (Select THREE.)

Accepted Answer

Multicollinearity between features. Multicollinearity occurs when two or more features in a dataset are highly correlated, meaning they contain redundant information. During exploratory data analysis (EDA), correlation matrices and variance inflation factor (VIF) calculations can reveal this issue, which can destabilize linear regression models and inflate coefficient standard errors.

Answer

High latency in API endpoints

Answer

Gradient vanishing in neural networks

Question 10

A data scientist is training a binary classifier on a dataset with 1 million rows and 500 features. The model uses XGBoost and achieves an AUC of 0.95 on the training set but only 0.72 on the test set. The scientist suspects overfitting. Which combination of hyperparameter adjustments is most likely to improve generalization?

Accepted Answer

Decrease 'max_depth' and increase 'min_child_weight'. Decreasing 'max_depth' reduces the complexity of individual trees, preventing the model from learning overly specific patterns in the training data. Increasing 'min_child_weight' forces the algorithm to require a higher sum of instance weights (hessian) before further partitioning, which acts as a regularization mechanism that discourages splits on noisy or sparse data. Together, these adjustments directly combat overfitting in XGBoost by limiting tree depth and requiring more evidence for splits, which improves generalization from the training AUC of 0.95 to a higher test AUC.

Answer

Increase 'max_depth' and decrease 'learning_rate'

Answer

Increase 'subsample' and decrease 'colsample_bytree'

Answer

Decrease 'gamma' and increase 'learning_rate'

Question 11

A machine learning engineer is deploying a model using SageMaker and wants to use automatic scaling for the endpoint based on the number of concurrent requests. The engineer has defined a scaling policy using the SageMakerVariantInvocationsPerInstance metric. However, the scaling is not triggering as expected. What could be the issue?

Accepted Answer

The scaling policy does not have a cooldown period configured, or the cooldown period is too long.. A missing or excessively long cooldown period can prevent the scaling policy from triggering. Cooldown periods (default 300 seconds) allow metrics to stabilize before initiating another scaling activity. Option A is incorrect because scheduled scaling actions are separate from dynamic scaling policies and are not required. Option C is incorrect because the SageMakerVariantInvocationsPerInstance metric is automatically published to CloudWatch. Option D is incorrect because this metric is specifically designed for automatic scaling.

Answer

A scheduled scaling action must be created first.

Answer

The metric must be published to CloudWatch manually.

Answer

The metric is not available for automatic scaling.

Question 12

A company is using SageMaker to train a model and wants to ensure that the training data is encrypted at rest and in transit, and that the trained model artifacts are also encrypted. Which THREE actions should the company take?

Accepted Answer

Specify a KMS key in the SageMaker training job configuration to encrypt the ML storage volume. Options A, B, and E are correct. A: Specifying a KMS key in the SageMaker training job configuration encrypts the ML storage volume used during training. B: Enabling SageMaker model encryption with a KMS key encrypts the model artifacts. E: Enabling S3 server-side encryption (SSE-KMS) on the training data bucket encrypts the data at rest in S3 and in transit (when SageMaker accesses it). Option C (VPC with no internet access) provides network isolation but not encryption. Option D (CloudTrail) is for auditing API calls, not encryption.

Answer

Configure the training job to run in a VPC with no internet access

Answer

Enable AWS CloudTrail to log all API calls

Question 13

A company is deploying a machine learning model using AWS Lambda for real-time inference. The model is a large ensemble model that takes approximately 500 MB of memory. The Lambda function is configured with 1024 MB of memory and a timeout of 15 seconds. The company observes that the function frequently times out during inference. The company wants to keep using Lambda for its serverless benefits. Which solution should the company implement to reduce inference time?

Accepted Answer

Increase the Lambda function memory to 3008 MB to provide more CPU resources.. Lambda has a maximum memory of 10,240 MB and a maximum timeout of 15 minutes. Increasing memory to 3008 MB gives more CPU power and reduces inference time. Option A is correct. Option B (SageMaker) moves away from serverless, which the company wants to keep. Option C (Step Functions) adds orchestration overhead and does not directly reduce inference time. Option D (ElastiCache) adds latency and cost and does not address the timeout issue.

Answer

Deploy the model on Amazon SageMaker hosting instead of Lambda.

Answer

Use AWS Step Functions to invoke the Lambda function asynchronously.

Answer

Use Amazon ElastiCache to cache model predictions and reduce computation.

Question 14

Which TWO statements about handling missing data during exploratory data analysis are correct? (Select TWO.)

Accepted Answer

Visualizing the pattern of missingness can help determine if data is missing at random.. Options B and C are correct. Visualizing the pattern of missingness helps determine if data is missing at random, which is a key EDA step. Understanding the missing data mechanism (MCAR, MAR, MNAR) is important for selecting an appropriate imputation strategy. Option A is incorrect because missing values should be addressed during EDA, not deferred to model training. Option D is incorrect because listwise deletion can introduce bias if data is not MCAR. Option E is incorrect because mean imputation reduces the variance of the imputed variable.

Answer

Missing values can be ignored during EDA and handled during model training.

Answer

Listwise deletion (removing rows with missing values) is always safe and unbiased.

Answer

Imputing missing values with the mean preserves the original variance.

Question 15

A data scientist is setting up an IAM role for an Amazon SageMaker training job. The policy shown is attached to the role. The training job fails with an access denied error when trying to read the training data from s3://my-bucket/training/data.csv. What is the most likely reason?

Accepted Answer

The bucket policy on my-bucket denies access to the IAM role. The IAM policy shown grants s3:GetObject access to the bucket and object, but the training job still fails with an access denied error. The most likely cause is that the bucket policy on my-bucket explicitly denies access to the IAM role, overriding the IAM policy's allow. In AWS, an explicit deny in a resource-based policy (bucket policy) takes precedence over any allow in an identity-based policy (IAM role policy), causing the access denied error despite the IAM policy appearing sufficient.

Answer

The training job is using a different IAM role

Answer

The IAM policy is missing the s3:ListBucket permission

Answer

The IAM role does not have permission to access the bucket location

Question 16

A company is using SageMaker to train a linear regression model on a dataset that fits into memory on a single instance. The training job is taking longer than expected. The data scientist wants to reduce training time without changing the algorithm. Which approach is most effective?

Accepted Answer

Use SageMaker's distributed training with multiple instances.. SageMaker's distributed training with multiple instances performs data parallelism, splitting the dataset across instances to train concurrently, which reduces training time for data that fits in memory. Option A is wrong because disabling automatic model tuning (hyperparameter tuning) does not speed up training; it only stops searching for optimal hyperparameters. Option B is wrong because while a larger instance with more vCPUs can help, distributed training scales better and is more cost-effective for this scenario. Option D is wrong because reducing the number of epochs would likely underfit the model, hurting accuracy.

Answer

Disable automatic model tuning.

Answer

Use a larger instance type with more vCPUs.

Answer

Reduce the number of epochs.

Question 17

A data scientist is setting up an IAM policy for a SageMaker notebook instance that needs to read and write data in the 'training/' folder of an S3 bucket, and also list objects in the bucket. Does the policy satisfy the requirements?

Accepted Answer

Yes, the policy correctly grants the required permissions.. The stem does not include the IAM policy, so the question cannot be definitively answered as written. However, assuming a typical policy that grants s3:GetObject and s3:PutObject on the training/ prefix and s3:ListBucket with a condition restricting the prefix to training/*, the policy meets the requirements. Therefore, option A is correct under that interpretation. Option B is wrong because s3:DeleteObject is not required for reading and writing. Option C is wrong because s3:GetObject is allowed on the training/ objects, not on the bucket itself. Option D is wrong because the condition on ListBucket is valid and correctly limits listing to the training/ prefix.

Answer

No, the policy must also include s3:DeleteObject for data cleaning.

Answer

No, the policy misses s3:GetObject for the bucket itself.

Answer

No, the condition on ListBucket is invalid.

Question 18

A company wants to deploy a machine learning model on Amazon SageMaker and needs to monitor the model's performance in production. Which TWO AWS services can be used to set up monitoring?

Accepted Answer

Amazon CloudWatch. Amazon CloudWatch is correct because it provides comprehensive monitoring for SageMaker endpoints, including metrics like latency, invocation counts, and error rates. It can also trigger alarms and dashboards for performance degradation. Amazon SageMaker Model Monitor is correct because it specifically detects data drift, feature attribution drift, and quality issues in production models by analyzing inference data against a baseline.

Answer

AWS X-Ray

Answer

Amazon Inspector

Answer

AWS Config

Question 19

Refer to the exhibit. An IAM policy is attached to a SageMaker notebook instance. Which action will the notebook be able to perform?

Accepted Answer

Invoke a SageMaker endpoint. The IAM policy attached to the SageMaker notebook instance grants only the `sagemaker:InvokeEndpoint` action. This action allows the notebook to send inference requests to a deployed SageMaker endpoint. No other SageMaker or S3 actions are permitted, so the notebook can only invoke the endpoint.

Answer

Create a training job

Answer

Create a model

Answer

Read data from S3

Question 20

A data scientist runs a SageMaker notebook and uses pandas to explore a dataset. The dataset contains 500,000 rows and 20 columns, including a 'timestamp' column. After loading the data into a DataFrame, the memory usage is unexpectedly high. What is the most likely cause?

Accepted Answer

The default data types inferred by pandas are unnecessarily large for the actual data ranges.. When pandas reads a CSV without explicit dtypes, it infers data types. For numeric columns, it defaults to int64 (8 bytes per value) or float64 (8 bytes per value), even if the actual values could fit in smaller types (e.g., int8, int16). With 500,000 rows and 20 columns, such large types significantly increase memory usage. Option A is wrong because pandas does not automatically create an index from the timestamp column; it assigns a default integer index. Option C is wrong because the entire dataset was loaded, not just a sample. Option D is wrong because the CSV is uncompressed; compression would not cause memory inflation after loading.

Answer

The DataFrame created an index column on the timestamp field, doubling memory usage.

Answer

The DataFrame only loaded a sample of the data, but the sample size was too large.

Answer

The CSV file was compressed, and pandas inflated it in memory.