Types of machine learning quiz

Question 1

What are the three main types of machine learning?

Accepted Answer

Supervised, unsupervised, reinforcement

Answer

Fast, slow, medium

Answer

Hot, cold, warm

Answer

Input, output, process

Question 2

In supervised learning, what does 'labeled data' mean?

Accepted Answer

Data with the correct answer attached to each example

Answer

Data that is colorful

Answer

Data stored on a label printer

Answer

Data that is always wrong

Question 3

Which type of machine learning is used when a model learns to sort emails into 'spam' and 'not spam' using past examples that are already marked?

Accepted Answer

Supervised learning

Answer

Unsupervised learning

Answer

Reinforcement learning

Answer

No learning

Question 4

What is unsupervised learning?

Accepted Answer

Finding patterns in data without labels

Answer

Learning only from a teacher

Answer

Memorizing the answer key

Answer

Learning from rewards only

Question 5

What is reinforcement learning, in simple words?

Accepted Answer

Learning by trial and error using rewards and penalties

Answer

Learning from labeled photos

Answer

Copying another program

Answer

Sorting data into folders

Question 6

A robot learns to walk by getting points when it moves forward and losing points when it falls. Which type of learning is this?

Accepted Answer

Reinforcement learning

Answer

Supervised learning

Answer

Unsupervised learning

Answer

Semi-supervised learning

Question 7

What is the difference between classification and regression in supervised learning?

Accepted Answer

Classification predicts categories; regression predicts numbers

Answer

Both predict only numbers

Answer

Classification predicts numbers; regression predicts categories

Answer

They are exactly the same

Question 8

What is clustering, and which type of machine learning does it belong to?

Accepted Answer

Grouping similar data without labels; unsupervised learning

Answer

Predicting house prices; supervised learning

Answer

Earning rewards; reinforcement learning

Answer

Labeling images by hand; supervised learning

Question 9

What is semi-supervised learning?

Accepted Answer

Using a little labeled data plus lots of unlabeled data

Answer

Using only labeled data

Answer

Using only rewards

Answer

Using no data at all

Question 10

Why is labeled data often expensive or hard to get?

Accepted Answer

Humans must manually mark each example, costing time and money

Answer

Computers refuse to read it

Answer

Labels are illegal to use

Answer

It is always free and instant

Question 11

Which type of learning would you use to group photos of animals when you don't know in advance which animals appear?

Accepted Answer

Unsupervised learning

Answer

Supervised learning

Answer

Reinforcement learning

Answer

Semi-supervised learning

Question 12

In reinforcement learning, what is an 'agent'?

Accepted Answer

The learner that takes actions and receives rewards

Answer

The dataset of labeled images

Answer

The list of correct answers

Answer

The computer's keyboard

Question 13

What is a 'reward' in reinforcement learning?

Accepted Answer

A signal telling the agent how good its action was

Answer

A type of labeled photo

Answer

A cluster of data points

Answer

The training dataset

Question 14

A model predicts tomorrow's temperature as a number like 31.5 degrees. Is this classification or regression?

Accepted Answer

Regression

Answer

Classification

Answer

Clustering

Answer

Reinforcement learning

Question 15

A model decides whether a photo shows a 'dog' or 'not a dog'. Is this classification or regression?

Accepted Answer

Classification

Answer

Regression

Answer

Clustering

Answer

Reinforcement learning

Question 16

What is the main goal of unsupervised learning compared to supervised learning?

Accepted Answer

To discover hidden patterns in unlabeled data

Answer

To memorize all the answers

Answer

To always earn rewards

Answer

To copy the teacher exactly

Question 17

Which type of machine learning powers a game-playing AI that improves by playing thousands of matches against itself?

Accepted Answer

Reinforcement learning

Answer

Supervised learning

Answer

Unsupervised learning

Answer

Semi-supervised learning

Question 18

What are 'features' in a machine learning dataset?

Accepted Answer

The input attributes describing each example

Answer

The reward signal

Answer

The final accuracy score

Answer

The name of the algorithm

Question 19

In supervised learning, what is the difference between training data and test data?

Accepted Answer

Training teaches the model; test checks it on unseen data

Answer

They are identical

Answer

Test data trains the model; training data checks it

Answer

Both are used only for rewards

Question 20

What is 'dimensionality reduction', and which learning type is it usually part of?

Accepted Answer

Reducing the number of features; unsupervised learning

Answer

Adding more rewards; reinforcement learning

Answer

Adding many labels; supervised learning

Answer

Deleting all the data; no learning

Question 21

Movie streaming apps suggest films similar to ones you liked. Which idea from unsupervised learning often helps here?

Accepted Answer

Clustering similar items or users

Answer

Earning rewards for each click

Answer

Manually labeling every movie

Answer

Predicting exact ticket prices

Question 22

Why can't we always use supervised learning for every problem?

Accepted Answer

Because labeled data is not always available

Answer

Because computers dislike supervised learning

Answer

Because it is illegal

Answer

Because it never works at all

Question 23

What is overfitting in supervised learning?

Accepted Answer

The model memorizes training data and fails on new data

Answer

The model is too simple to learn anything

Answer

The model uses too few features

Answer

The model earns too many rewards

Question 24

In the dog-training analogy for reinforcement learning, what plays the role of the 'reward'?

Accepted Answer

The treat given for good behavior

Answer

The dog's leash

Answer

The training dataset

Answer

The list of labels

Question 25

Which type of machine learning is most suitable for detecting unusual or fraudulent bank transactions when most data is unlabeled?

Accepted Answer

Unsupervised learning (anomaly detection)

Answer

Pure supervised learning

Answer

Reinforcement learning

Answer

No learning is possible

Question 26

What is an 'environment' in reinforcement learning?

Accepted Answer

The world the agent interacts with and gets feedback from

Answer

The labeled training set

Answer

The accuracy score

Answer

The list of correct answers

Question 27

A self-driving car learns to drive safely by getting penalties for crashes and rewards for smooth, safe driving. Which learning type is this?

Accepted Answer

Reinforcement learning

Answer

Supervised learning

Answer

Unsupervised learning

Answer

Semi-supervised learning

Question 28

How is semi-supervised learning different from supervised learning?

Accepted Answer

Semi-supervised mixes labeled and unlabeled data

Answer

They are exactly the same

Answer

Semi-supervised uses only rewards

Answer

Supervised uses no labels at all

Question 29

Why is testing a supervised model on new, unseen data important?

Accepted Answer

It checks if the model generalizes to new examples

Answer

It makes the model train faster

Answer

It deletes the labels

Answer

It earns extra rewards

Question 30

Which simple Python library is commonly used to try supervised and unsupervised learning?

Accepted Answer

scikit-learn (sklearn)

Answer

calculator

Answer

notepad

Answer

paint

Test yourself