Computer Science
Grade 10
20 min
Supervised Learning
Supervised Learning
Tutorial Preview
1
Introduction & Learning Objectives
Learning Objectives
Define Supervised Learning and explain its core purpose.
Differentiate between classification and regression tasks with examples.
Identify features and labels in a given dataset.
Describe the high-level process of training a model using labeled data.
Explain how a trained model makes a prediction on new, unseen data.
List at least three real-world applications of Supervised Learning.
How does your email app know which messages are junk, and how does a streaming service know which movie you'll love next? 🤔 Let's learn about the 'teacher' that helps computers make these smart guesses!
This lesson introduces Supervised Learning, a fundamental type of machine learning where we teach a computer to make predictions by showing it lots of ex...
2
Key Concepts & Vocabulary
TermDefinitionExample
Supervised LearningA type of machine learning where an algorithm learns from a dataset that has been labeled with the correct answers. It's like learning with a teacher or an answer key.Showing a computer 1,000 pictures of cats labeled 'cat' and 1,000 pictures of dogs labeled 'dog' so it can learn to identify them on its own.
Training DataThe initial dataset of labeled examples used to teach the machine learning model. The model studies this data to find patterns.A spreadsheet with 500 rows, where each row has the square footage of a house, number of bedrooms, and its final selling price (the label).
FeaturesThe input variables or attributes used by the model to make a prediction. These are the pieces of information you already know.To predic...
3
Core Syntax & Patterns
The Supervised Learning Workflow
1. Gather Labeled Data -> 2. Split Data (Train/Test) -> 3. Train the Model -> 4. Evaluate the Model -> 5. Make Predictions
This is the fundamental step-by-step process for any supervised learning project. You start with data that has answers, teach a model with most of it, check its performance with the rest, and then use the trained model on new data.
The Core Prediction Pattern
Model(New Features) -> Predicted Label
Once a model is trained, its job is to take a new set of features (for which you don't know the answer) and produce a predicted label. This is the 'inference' or 'prediction' phase.
Choosing Your Task: Classification vs. Regression
IF the label is a category (e.g., 'Yes'/&#...
4 more steps in this tutorial
Sign up free to access the complete tutorial with worked examples and practice.
Sign Up Free to ContinueSample Practice Questions
Challenging
A model is trained to predict house prices (a regression task), but the training data only contains small, one-bedroom apartments priced under $100,000. If you ask this model to predict the price of a five-bedroom mansion, what is the most likely outcome based on the 'Garbage In, Garbage Out' principle?
A.The model will likely make a very inaccurate prediction because it has never seen data for a house that large.
B.The model will accurately predict the price because regression can handle any numbers.
C.The model will crash because the input features are too different.
D.The model will predict a price very close to $100,000, the maximum it has seen.
Challenging
Imagine building a model to classify images as 'Cat' or 'Dog'. The 'features' are the pixel values of an image, and the 'label' is the word 'Cat' or 'Dog'. How does the model use the labeled training data to learn the difference?
A.It memorizes every single training image and its label.
B.It learns to sort the images alphabetically by their labels.
C.It ignores the pixel values and focuses only on the labels.
D.It identifies patterns in pixel values that are consistently associated with the 'Cat' label and different patterns associated with the 'Dog' label.
Challenging
In the 'Will a Student Pass?' example, a model is trained on data that only includes students who studied for 1 hour (and failed) or 15 hours (and passed). Why might this model perform poorly when predicting the outcome for a new student who studied for 8 hours?
A.Because 8 hours is a feature, not a label.
B.Because the model has no information about the middle-range of study hours and doesn't know the pattern in that region.
C.Because the model can only predict 'Pass' or 'Fail', which is a classification task.
D.Because the training data is not labeled correctly.
Want to practice and check your answers?
Sign up to access all questions with instant feedback, explanations, and progress tracking.
Start Practicing Free