NEW
Font size
Worksheets12B,D,E final revision
Total questions: 50
Worksheet time: 25mins
What is the main goal of Artificial Intelligence (AI)?
To store large amounts of data
To help machines think and make decisions like humans
To make computers larger and faster
To reduce the number of computer programs needed
Machine learning allows computers to:
Memorize all data without understanding
Learn from data and make predictions
Write code automatically
Never make mistakes
Which of the following is a dataset?
A single number
A collection of data used to train or test a model
A programming language
A website
A feature is:
A value used to make predictions
A final prediction
A programming function
The label
A label is:
The dataset’s title
The value the model is trying to predict
The number of features
A programming command
Supervised learning uses:
Labeled data
No labels
Random guessing
Predictions only
Unsupervised learning uses:
Labeled data
Data without labels
Only numbers
Only text
Classification models:
Predict categories
Predict numerical values
Delete features
Create datasets
Which is a classification task?
Predicting the number of apples in a basket
Deciding whether a photo shows a cat or dog
Calculating the sum of numbers
Measuring temperature
Regression models:
Group data into categories
Predict numerical values
Ignore features
Make random predictions
Training a model means:
It learns patterns from data
Testing it on new data
Deleting data
Random predictions
Testing a model means:
Training it further
Checking its accuracy on new data
Adding features
Removing labels
Bias occurs when:
Data is balanced
Data is incomplete or unrepresentative
Model predicts correctly
Features are too many
Example of supervised learning:
Grouping customers without labels
Classifying emails as spam or not
Random number generation
Sorting numbers alphabetically
Example of unsupervised learning:
Predicting house prices
Grouping customers by buying behavior without labels
Classifying emails
Predicting exam scores
Patterns in data help models:
Memorize the dataset
Make predictions
Randomize outputs
Delete features
Multiple features in a dataset:
Confuse the model
Give more information to improve predictions
Increase file size unnecessarily
Slow down training
Labeling data provides:
Correct answers for the model
Extra features
Random guesses
Larger datasets
Real-life example of AI:
Chess-playing program
Visual impairment app
Email spam filter
All of the above
A good dataset should be:
Small and incomplete
Large, diverse, and representative
Random without relevance
Only one type of example
AI can help humans by:
Automating repetitive tasks
Supporting decision-making
Predicting trends from data
All of the above
Features can be:
Numbers, text, or categories
Only numbers
Only images
Only text
A model’s prediction depends on:
Features and patterns in data
Random numbers
The size of the computer
The number of labels only
Patterns in data are important because:
They reduce dataset size
They allow models to learn and predict
They make data more colorful
They increase memory usage
Labeling is necessary for:
Unsupervised learning
Supervised learning
Random guessing
Testing only
AI can fail if:
Data is biased or incomplete
Model is simple
Features are missing
All of the above
Evaluating a model helps to:
Increase features
Improve accuracy and detect bias
Reduce dataset size
Delete labels
Bias can cause:
Fair predictions
Unfair or inaccurate predictions
More features
Larger datasets
A dataset with missing values can:
Improve model accuracy
Reduce model accuracy
Increase bias automatically
Make the model unsupervised
AI examples in daily life include:
Voice assistants
Self-driving cars
Recommendation systems
All of the above
Which of these is a common step in machine learning?
Training
Testing
Evaluating
All of the above
A feature is important because it:
Determines what the model can learn from
Is always the predicted value
Reduces dataset size
Is optional
Which of these is supervised learning?
Grouping customers by behavior
Predicting stock prices from historical data
Both A and B
None of the above
Which of these is unsupervised learning?
Clustering customers based on purchasing habits
Predicting tomorrow’s temperature
Classifying emails as spam
Counting students in class
A classification model predicts:
Categories
Numbers
Text only
Images only
A regression model predicts:
Categories
Numerical values
Labels only
Features only
Patterns in data allow models to:
Memorize data exactly
Make predictions on new data
Ignore new inputs
Delete labels
Evaluating a model helps to:
Detect errors or bias
Delete features
Memorize the dataset
Make predictions random
An example of AI improving lives is:
A recommendation system for movies
A voice assistant like Siri
Self-driving car software
All of the above
Bias in AI can be caused by:
Incomplete data
Data that is not diverse
Ignoring certain features
All of the above
A model card is used to:
Represent information about a trained model
Delete a dataset
Add features automatically
Train the model faster
Supervised learning requires:
Labeled data
Unlabeled data
Random guessing
No data
Unsupervised learning requires:
Labeled data
Unlabeled data
Random guesses
Predefined predictions
Classification vs Regression:
Both predict categories
Classification predicts categories, regression predicts numbers
Regression predicts categories, classification predicts numbers
Both predict numbers
Features can be:
Age, height, color, or category
Only numbers
Only letters
Only images
A model makes predictions based on:
Patterns in training data
Random numbers
File size
Number of labels only
Which is an example of classification?
Deciding whether a customer will buy a product or not
Predicting the total sales amount
Counting items in inventory
Measuring temperature
Which is an example of regression?
Predicting tomorrow’s temperature
Sorting fruits into categories
Detecting spam emails
Grouping customers
Patterns help a model because:
They allow predictions on new data
They increase file size
They reduce accuracy
They delete features
Evaluating a model ensures:
It works accurately and fairly
It deletes data automatically
It adds random features
It ignores bias
