Finding and using datasets quiz

Question 1

What is a dataset?

Accepted Answer

A collection of related data organized for analysis

Answer

A single data point stored in a variable

Answer

A computer program that processes numbers

Answer

A chart showing statistics

Question 2

What does 'structured data' mean?

Accepted Answer

Data organized in rows and columns, like a spreadsheet

Answer

Data stored on a USB drive

Answer

Images and audio files

Answer

Encrypted private data

Question 3

What does CSV stand for, and why is it a popular format for sharing datasets?

Accepted Answer

Comma-Separated Values

Answer

Computed Structured Variables

Answer

Centralized Storage Volume

Answer

Coded Statistical Values

Question 4

Which of the following is a well-known free platform where data scientists share machine-learning datasets and compete in challenges?

Accepted Answer

Kaggle

Answer

GitHub Actions

Answer

Stack Overflow

Answer

Replit

Question 5

In a labeled dataset used for machine learning, what does a 'label' represent?

Accepted Answer

The correct answer or outcome the model should predict

Answer

The name given to a column in a spreadsheet

Answer

The source website where the data was downloaded

Answer

A short description written by the researcher

Question 6

What is a 'feature' in a machine-learning dataset?

Accepted Answer

A measurable input property used to make predictions

Answer

A decorative element added to a chart

Answer

The final prediction made by the model

Answer

A special type of CSV file

Question 7

Why are missing values in a dataset a problem for machine learning?

Accepted Answer

Algorithms may crash or give wrong results when data is incomplete

Answer

Missing values make the file size larger

Answer

Missing values always mean the data was collected illegally

Answer

Missing values slow down the internet connection

Question 8

What is a train/test split, and why do data scientists use it?

Accepted Answer

Dividing data into one part for teaching the model and one part for evaluating it

Answer

Splitting a dataset by deleting every other row

Answer

Separating numerical columns from text columns

Answer

Creating two backup copies of the dataset

Question 9

What does 'bias in a dataset' mean, and why is it harmful for AI systems?

Accepted Answer

When data unfairly over- or under-represents certain groups, leading to unfair AI

Answer

When the dataset file is corrupted and cannot be opened

Answer

When the dataset is too large to download quickly

Answer

When labels are stored in a different language

Question 10

Why is it important to protect personal data in a dataset?

Accepted Answer

Sharing personal data without permission can harm people and break privacy laws

Answer

Personal data makes files too big to process

Answer

Personal data is never useful for training AI models

Answer

Protecting data automatically makes the model more accurate

Question 11

The UCI Machine Learning Repository is a famous source of free datasets. Which organization hosts it?

Accepted Answer

University of California, Irvine

Answer

United Nations Computing Institute

Answer

US Central Intelligence Agency

Answer

United Computing Industry Consortium

Question 12

What is 'open data,' and where can you find government open data?

Accepted Answer

Freely accessible data that anyone can use, often published by governments

Answer

Data that has been unlocked by hackers

Answer

Data stored without any encryption

Answer

Data collected only from public parks

Question 13

Why does the license of a dataset matter before you use it in a project?

Accepted Answer

The license defines what you are legally allowed to do with the dataset

Answer

The license tells you the programming language used to create the dataset

Answer

The license determines how fast the dataset downloads

Answer

Licenses only matter for datasets with more than 10,000 rows

Question 14

What is 'overfitting,' and how does having too little training data cause it?

Accepted Answer

When a model memorizes training examples instead of learning general patterns

Answer

When a model performs perfectly on new data but poorly on training data

Answer

When a dataset file is too large to fit on a hard drive

Answer

When two datasets are accidentally merged together

Question 15

What is data augmentation, and which type of dataset benefits most from it?

Accepted Answer

Creating new training examples by modifying existing data, especially useful for images

Answer

Adding new statistical columns to the dataset

Answer

Downloading a second copy of the same dataset for backup

Answer

Converting all data types to floating-point numbers

Question 16

What is an imbalanced dataset, and why can it mislead a machine-learning model?

Accepted Answer

A dataset where one class has far more examples than another, potentially misleading the model

Answer

A dataset where columns have different data types

Answer

A dataset split unevenly between training and testing

Answer

A dataset stored on two different servers

Question 17

Ice cream sales and drowning incidents both increase in summer. Does this mean eating ice cream causes drowning?

Accepted Answer

No — both are caused by hot weather; correlation does not imply causation

Answer

Yes, because the data clearly shows both rising together

Answer

Yes, but only if the dataset has more than 1,000 rows

Answer

No — the data must be wrong because the relationship is too strong

Question 18

What is the MNIST dataset, and why is it famous in the history of AI?

Accepted Answer

A dataset of 70,000 labeled handwritten digit images used to benchmark classification algorithms

Answer

A dataset of medical scans used to detect tumors

Answer

A collection of news articles labeled by topic

Answer

A government dataset of census population records

Question 19

What does the following Python code do when you first load a new dataset?

Accepted Answer

Loads a CSV file into a table and displays the first five rows

Answer

Downloads a file from the internet named animals.csv

Answer

Counts how many animals are in the file

Answer

Saves the DataFrame back to a new CSV file

Question 20

What is an outlier in a dataset, and how should a data scientist handle it?

Accepted Answer

A data point that is very different from the rest of the data

Answer

A column that contains only text values

Answer

A row that has been duplicated in the dataset

Answer

A feature that is perfectly correlated with the label

Question 21

Which of the following is an example of categorical data?

Accepted Answer

Eye color of students (Brown, Blue, Green)

Answer

Temperature recorded in degrees Celsius

Answer

Number of goals scored in a match

Answer

Distance run in a marathon

Question 22

What is data anonymization, and how does it protect people in AI projects?

Accepted Answer

Removing or disguising identifying information to protect individuals' privacy

Answer

Encrypting a dataset so only one person can read it

Answer

Deleting all rows that contain numbers

Answer

Adding fake rows to confuse hackers

Question 23

A dataset has no labels at all. Which type of machine learning would you most likely use with it?

Accepted Answer

Unsupervised learning, where the model finds patterns without labels

Answer

Supervised learning, because labels can be added later

Answer

Reinforcement learning, because the agent earns rewards instead

Answer

Deep learning, which never requires any labels

Question 24

Why does having a larger and more diverse dataset generally lead to a better machine-learning model?

Accepted Answer

More diverse examples help the model generalize and handle a wider variety of cases

Answer

Larger files take longer to process, giving the model more thinking time

Answer

Large datasets always make the model run faster after training

Answer

Diversity in data removes the need for any data cleaning

Question 25

Is it ethical to automatically scrape (collect) data from a website without the site owner's permission?

Accepted Answer

No — scraping without permission can violate terms of service, laws, and ethical norms

Answer

Yes, because all data on the internet is public and free to use

Answer

Yes, as long as the data is only used for a school project

Answer

It is fine as long as the website does not display a copyright symbol

Question 26

What is an API, and how can it help you gather data for a machine-learning project?

Accepted Answer

A programming interface that lets you request structured data from online services

Answer

A type of encrypted database that stores images securely

Answer

A chart automatically generated from a CSV file

Answer

A government law that regulates AI datasets

Question 27

Why is it important to normalize (scale) numerical features before training many machine-learning models?

Accepted Answer

Normalization brings all features to a similar scale so no single feature unfairly dominates

Answer

Normalization automatically removes missing values from the dataset

Answer

Normalization converts text columns into numbers

Answer

Normalization speeds up the file download from Kaggle

Question 28

What is cross-validation, and why is it more reliable than a single train/test split?

Accepted Answer

Training on multiple folds and averaging results so every data point is tested at least once

Answer

Testing the model on the same data it was trained on, repeated many times

Answer

Splitting the dataset into more than three parts to increase its size

Answer

Checking that the CSV file has no formatting errors before training starts

Question 29

ImageNet is a benchmark dataset containing over 14 million images. What is it primarily used for in AI?

Accepted Answer

Training and evaluating image classification and object-detection models

Answer

Training models to understand written text in many languages

Answer

Testing speech-recognition systems across different accents

Answer

Storing satellite photographs for government mapping agencies

Question 30

What do the following two lines of Python code tell you about a dataset?

Accepted Answer

Shows the dataset dimensions and counts missing values per column

Answer

Deletes null values and resizes the dataset automatically

Answer

Displays the first and last rows of the dataset

Answer

Saves the dataset shape to a text file

Test yourself