MLS-C01 Exploratory Data Analysis Practice Test 2 — 15 Questions

Question 1

A data scientist runs the following AWS CLI command:
aws s3api head-object --bucket my-bucket --key data.csv

The output is:
{
    "AcceptRanges": "bytes",
    "LastModified": "2021-08-21T12:00:00+00:00",
    "ContentLength": 1048576,
    "ETag": "\"d41d8cd98f00b204e9800998ecf8427e\"",
    "ContentType": "text/csv",
    "Metadata": {}
}

What can be concluded from the output?

Accepted Answer

The ETag can be used for integrity checking.. The ETag can be used for integrity checking, as it is an MD5 hash of the object content. Option B is wrong because ContentLength 1048576 corresponds to 1 MB, not 1 GB. Option C is wrong because S3 versioning is indicated by VersionId, not ETag. Option D is wrong because the output does not provide any information about preprocessing; the presence of metadata like 'preprocessed' is not shown.

Answer

The file is 1 GB in size.

Answer

The object has S3 versioning enabled.

Answer

The file has not been preprocessed.

Question 2

A data scientist is troubleshooting access to an S3 bucket. The following IAM policy is attached to their role. What is the likely result when they try to list objects in the 'confidential' folder?

```json
{
  "Version": "2012-10-17",
  "Statement": [
    {
      "Effect": "Allow",
      "Action": "s3:ListBucket",
      "Resource": "arn:aws:s3:::example-bucket"
    },
    {
      "Effect": "Deny",
      "Action": "s3:*",
      "Resource": "arn:aws:s3:::example-bucket/confidential/*",
      "Condition": {
        "StringNotEquals": {
          "aws:SourceVpce": "vpce-12345678"
        }
      }
    }
  ]
}
```

Accepted Answer

Access is denied if the request does not originate from the specified VPC endpoint.. The Deny statement uses a StringNotEquals condition on aws:SourceVpce, meaning the deny only fires when the request does NOT come from vpce-12345678. Because the Allow statement grants s3:ListBucket on the bucket, requests originating from that VPC endpoint are permitted, while requests from anywhere else are explicitly denied. Explicit Deny always overrides Allow in IAM evaluation.

Answer

Access is allowed because the Allow statement grants s3:ListBucket.

Answer

Access is denied unconditionally.

Answer

Access is allowed only if the request uses HTTPS.

Question 3

A data scientist is reviewing a dataset and notices that the distribution of a numerical feature is heavily right-skewed with a long tail. Which visualization is most appropriate to assess the distribution?

Accepted Answer

Histogram with a logarithmic scale on the x-axis. A histogram with a logarithmic scale on the x-axis is most appropriate for a heavily right-skewed distribution with a long tail because it compresses the large values and expands the small values, making the shape of the distribution visible across several orders of magnitude. A standard histogram would show most data in the first bin and a long flat tail, obscuring the distribution's form. The log scale reveals whether the tail follows a power law or exponential decay.

Answer

Box plot

Answer

Line chart

Answer

Scatter plot

Question 4

A data scientist is analyzing a dataset with missing values in several features. The dataset is large (10 million rows) and stored in an S3 bucket as CSV files. The scientist wants to use AWS Glue to catalog the data and then use Amazon Athena to query it. However, the missing values are causing errors in downstream machine learning models. Which approach should the scientist take to handle missing values during exploratory data analysis?

Accepted Answer

Use AWS Glue ETL jobs with a custom transformation script that uses the AWS Glue library to drop or impute missing values before writing to a new dataset.. AWS Glue ETL jobs can process large datasets (10 million rows) in a distributed Spark environment and apply custom transformations to drop or impute missing values, writing a clean dataset back to S3 for Athena and downstream ML. This is the most scalable and production-ready approach for handling missing values at scale before analysis.

Answer

Use Amazon SageMaker Data Wrangler to create a data flow that imputes missing values and export the transformed dataset to S3.

Answer

Use Amazon Redshift Spectrum with an external table to query the data and use SQL COALESCE to handle missing values on the fly.

Answer

Use Amazon Athena to run SQL queries that impute missing values and write the results to a new table.

Question 5

A machine learning engineer is performing exploratory data analysis on a dataset containing customer transaction records. The dataset includes a column 'transaction_date' with timestamps. The engineer wants to derive features such as day of the week, hour, and month for modeling. Which AWS service can be used directly to extract these features without writing custom code?

Accepted Answer

AWS Glue ETL with built-in timestamp transforms. AWS Glue ETL provides built-in transforms like `ExtractTimestamp` that can parse timestamps and extract date/time components (e.g., day of week, hour, month) without writing custom code. Option B is wrong because Amazon Athena requires writing SQL queries to extract date parts, which constitutes custom code. Option C is wrong because Amazon QuickSight is a BI visualization tool, not designed for feature engineering. Option D is wrong because Amazon SageMaker Data Wrangler, while offering visual transformations, requires an active SageMaker Studio environment and is not a serverless ETL service like AWS Glue.

Answer

Amazon Athena with SQL date functions

Answer

Amazon QuickSight

Answer

Amazon SageMaker Data Wrangler

Question 6

A machine learning engineer is analyzing a dataset with a large number of features (p >> n). The engineer suspects that many features are irrelevant. Which THREE methods are suitable for feature selection during exploratory data analysis? (Choose THREE.)

Accepted Answer

Remove features with variance below a threshold (e.g., <0.01). In a high-dimensional dataset (p >> n), feature selection is crucial. Option B (Variance Threshold) is suitable because features with low variance (e.g., <0.01) are likely to be constant or near-constant and thus uninformative. Option C (removing features with high pairwise correlation >0.95) helps reduce redundancy and multicollinearity. Option D (mutual information) is a filter method that measures dependency between each feature and the target, allowing selection of the most relevant features. Option A (Lasso regression) is a modeling method that can be used for feature selection but is not typically used during EDA; it is a supervised learning technique. Option E (PCA) is a dimensionality reduction technique that creates new components, not feature selection. Therefore, the correct answers are B, C, D.

Answer

Fit a Lasso regression model and select features with non-zero coefficients

Answer

Apply Principal Component Analysis (PCA) and select top components

Question 7

During exploratory data analysis on a dataset with 1 million rows, a data scientist notices that the distribution of the target variable is highly imbalanced (99% class A, 1% class B). Which technique should be applied to address this imbalance before model training?

Accepted Answer

Use SMOTE to generate synthetic samples for the minority class. SMOTE (Synthetic Minority Oversampling Technique) generates synthetic samples for the minority class, balancing the dataset. Option A is wrong because random undersampling can discard important data. Option B is wrong because scaling does not address imbalance. Option C is wrong because PCA does not fix imbalance.

Answer

Randomly undersample the majority class to match the minority class size

Answer

Apply standard scaling to all features

Answer

Use PCA to reduce dimensionality and oversample in principal component space

Question 8

A company stores customer transaction data in Amazon S3. A data scientist needs to perform exploratory data analysis using Amazon SageMaker. The dataset is 500 GB in CSV format. Which approach is most cost-effective and time-efficient for initial data profiling?

Accepted Answer

Use Amazon S3 Select to sample rows directly from S3. For initial exploratory data profiling of a 500 GB CSV dataset, Amazon S3 Select is the most cost-effective and time-efficient approach because it lets you run SQL-like queries to sample and filter rows directly from S3 without loading the full dataset. You pay only for the data scanned, and you avoid the cost and time of provisioning a large notebook instance or running ETL. This is ideal for a first look at schema, distributions, and anomalies.

Answer

Load the entire dataset into a SageMaker notebook instance and use pandas

Answer

Convert the data to Parquet format and then use Athena to query

Answer

Use AWS Glue ETL to transform the data and then analyze in Athena

Question 9

A data scientist is performing EDA on a dataset of customer churn. The dataset includes a categorical feature 'Region' with 100 unique values. What is the best way to encode this feature for a tree-based model?

Accepted Answer

Use the feature as a categorical variable directly in the tree-based model. Tree-based models (decision trees, random forests, gradient boosting) can natively handle categorical variables by evaluating splits on category subsets, so passing 'Region' as a categorical feature directly preserves the full information without imposing an artificial ordinal relationship. This avoids the dimensionality explosion of one-hot encoding (100 new columns) and the false ordering introduced by label encoding.

Answer

Replace each category with its frequency in the dataset

Answer

Label encode the feature (assign integers 0-99)

Answer

One-hot encode the feature

Question 10

Which TWO of the following are appropriate techniques for detecting outliers in a univariate continuous dataset? (Select TWO.)

Accepted Answer

Z-score method. The Z-score method (A) is appropriate because it standardizes each value as (x − μ)/σ and flags points whose absolute Z-score exceeds a threshold such as 3, directly identifying univariate outliers in a continuous variable. The IQR method (B) is also appropriate: it computes Q1 and Q3, defines the fences Q1 − 1.5·IQR and Q3 + 1.5·IQR (with IQR = Q3 − Q1), and labels values outside those fences as outliers, which is a robust univariate technique. Box plot visualization (C) is not a detection technique per se—it merely displays the same IQR-based summary graphically, so it is descriptive rather than an analytical method. Pearson correlation coefficient (D) measures linear association between two variables, so it is irrelevant to finding outliers in a single continuous variable. K-means clustering (E) is a multivariate partitioning algorithm and does not by itself define or detect univariate outliers.

Answer

Box plot visualization

Answer

Pearson correlation coefficient

Answer

K-means clustering

Question 11

A data scientist runs the above AWS CLI command. What does the command do?

Accepted Answer

It lists objects larger than 1,000,000 bytes under the data/ prefix.. The AWS CLI command `aws s3api list-objects-v2 --bucket my-bucket --prefix data/ --query 'Contents[?Size > `1000000`].[Key]'` uses the `--query` parameter with a JMESPath expression to filter the `Contents` array. The condition `Size > `1000000`` selects objects whose size in bytes exceeds 1,000,000 bytes (not 1 MB, which is 1,048,576 bytes). The projection `.[Key]` then extracts only the `Key` field for each matching object, so the output is a list of object keys (names) that satisfy the size condition. Thus, it lists objects larger than 1,000,000 bytes under the `data/` prefix.

Answer

It counts the number of objects larger than 1 MB.

Answer

It lists objects created after January 2023.

Answer

It lists objects larger than 1 MB in size.

Question 12

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 policy as described grants read and write access to the 'training/' folder (via s3:GetObject and s3:PutObject on the training/* resource) and list access on the bucket (via s3:ListBucket with a prefix condition). These permissions exactly match the stated requirements, so the policy is correct. No additional permissions are needed for the described use case.

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 13

A data scientist receives the above error during model training. What is the most likely cause?

Accepted Answer

The training data contains missing or infinite values.. The error message indicates that the training data contains missing (NaN) or infinite values, which causes the loss function to become NaN. This is a common issue when data has not been properly cleaned. Option B is wrong because a high learning rate typically leads to divergence or instability, not NaN values due to data issues. Option C is wrong because an incorrect data format would result in a parsing error, not a NaN loss. Option D is wrong because insufficient memory leads to an out-of-memory error, not NaN values.

Answer

The learning rate is too high.

Answer

The data format is incorrect; expected CSV but received JSON.

Answer

The instance type lacks sufficient memory.

Question 14

A data scientist is performing EDA on a large dataset (10 TB) stored in S3. They need to compute summary statistics for each column. Which approach is most cost-effective and efficient?

Accepted Answer

Use Amazon Athena with SQL queries. Amazon Athena is a serverless query service that allows you to run SQL queries directly on data stored in S3, charging only for the data scanned per query. This makes it highly cost-effective and efficient for computing summary statistics on large datasets without needing to provision or manage infrastructure. Option A is wrong because AWS Glue ETL jobs with PySpark have startup overhead and are better suited for complex data transformation tasks rather than simple ad-hoc analysis. Option C is wrong because downloading 10 TB to a SageMaker Studio notebook incurs high data transfer costs and requires substantial local storage, which is neither efficient nor cost-effective. Option D is wrong because launching an Amazon EMR cluster involves provisioning and managing compute resources, leading to higher costs and complexity for a straightforward statistical analysis that Athena can handle more simply.

Answer

Use an AWS Glue ETL job with PySpark to compute statistics

Answer

Download the dataset to an Amazon SageMaker Studio notebook and use pandas

Answer

Launch an Amazon EMR cluster and use Spark SQL

Question 15

A data engineer is exploring a dataset with a timestamp column and wants to resample the data to a consistent 1-hour frequency. The data is irregularly spaced. Which approach is most efficient using AWS services?

Accepted Answer

Use Amazon SageMaker Processing with a custom script. Amazon SageMaker Processing jobs allow custom scripts (e.g., using pandas resample) to handle irregular time series, and they are fully managed. Option A is wrong because Amazon EMR with Spark requires cluster management and is more complex for simple resampling. Option B is wrong because AWS Glue with built-in transforms is more suited for batch ETL but may be overkill for this task. Option C is wrong because Amazon Athena with SQL window functions is a query engine and cannot resample easily.

Answer

Use Amazon EMR with Spark

Answer

Use AWS Glue with built-in transforms

Answer

Use Amazon Athena with SQL window functions