Machine Learning Fundamentals

Active learning works because machine learning concepts become concrete when students manipulate real data. When students sort images, cluster cards, or test datasets, they see how algorithms adjust parameters based on patterns, making abstract functions visible. This hands-on approach builds lasting understanding by connecting mathematical ideas to tangible examples.

• Year 9
• Computing
• 4 activities
• 25 min–40 min

Learning Objectives

1. Classify machine learning algorithms as supervised or unsupervised based on provided examples of input data and desired output.
2. Analyze the impact of data quantity and quality on the accuracy of a simple machine learning model using a provided simulation or dataset.
3. Explain the fundamental process by which a machine learning model adjusts its parameters during training using a chosen analogy.
4. Compare and contrast the use cases for supervised and unsupervised learning in real-world applications.

Flipped Classroom

⏱ 35 min·Small Groups

Demo: Supervised Image Classifier

Provide printed animal images; students label half as training data and sort the rest as test data. Groups discuss matches and 'retrain' by adding more examples. Record accuracy before and after.

Prepare & details

Explain how a machine 'learns' from data without explicit programming.

Facilitation Tip: During the Demo: Supervised Image Classifier, show students the exact parameters the algorithm adjusts so they connect math to the visual output.

Setup: Standard classroom, flexible for group activities during class

Materials: Pre-class content (video/reading with guiding questions), Readiness check or entrance ticket, In-class application activity, Reflection journal

• Understand
• Apply
• Analyze
• Self-Management
• Self-Awareness

Flipped Classroom

⏱ 25 min·Pairs

Hands-on: Unsupervised Clustering

Give students unlabeled data cards with customer purchase traits. In pairs, they group cards into clusters without prior labels, then compare to a 'model' output. Reflect on patterns found.

Prepare & details

Differentiate between supervised and unsupervised learning with simple examples.

Facilitation Tip: During Hands-on: Unsupervised Clustering, ask groups to compare their clusters to a peer’s, highlighting how different starting points affect results.

Setup: Standard classroom, flexible for group activities during class

Materials: Pre-class content (video/reading with guiding questions), Readiness check or entrance ticket, In-class application activity, Reflection journal

• Understand
• Apply
• Analyze
• Self-Management
• Self-Awareness

Timeline Challenge

⏱ 40 min·Small Groups

Timeline Challenge: Data Quality Impact

Distribute biased and balanced datasets for predicting fruit ripeness. Small groups train simple paper models, test predictions, and swap datasets to observe performance drops. Chart results class-wide.

Prepare & details

Predict how the quality and quantity of training data impact a machine learning model's performance.

Facilitation Tip: During the Challenge: Data Quality Impact, provide a dataset with duplicates to demonstrate how noise disrupts pattern detection.

Setup: Long wall or floor space for timeline construction

Materials: Event cards with dates and descriptions, Timeline base (tape or long paper), Connection arrows/string, Debate prompt cards

• Remember
• Understand
• Analyze
• Self-Management
• Relationship Skills

Flipped Classroom

⏱ 30 min·Whole Class

Whole Class: Prediction Relay

Project a simple ML flowchart; teams relay to input training data examples verbally, predict outputs, and vote on model improvements. Adjust based on class feedback.

Prepare & details

Explain how a machine 'learns' from data without explicit programming.

Facilitation Tip: During the Prediction Relay, rotate roles so every student experiences both predictor and verifier to reinforce feedback loops.

Setup: Standard classroom, flexible for group activities during class

Materials: Pre-class content (video/reading with guiding questions), Readiness check or entrance ticket, In-class application activity, Reflection journal

• Understand
• Apply
• Analyze
• Self-Management
• Self-Awareness

Teaching This Topic

Teachers should focus on the role of data first, then introduce algorithms as tools that optimize based on examples. Avoid starting with code or complex math; instead, use sorting, grouping, and labeling tasks to build intuition. Research shows that students grasp supervised learning faster when they physically tag data, while unsupervised learning clicks when they see how grouping emerges without prior labels. Keep explanations grounded in concrete examples before abstracting.

What to Expect

Successful learning looks like students correctly labeling data for supervised tasks, identifying groupings in unsupervised sets, and articulating why data quality matters. They should explain the difference between labeled and unlabeled data and justify their choices with evidence from their activities.

Watch Out for These Misconceptions

Common Misconception

During Demo: Supervised Image Classifier, listen for students saying the computer 'understands' cats like humans do.

What to Teach Instead

Redirect by asking students to trace the algorithm’s steps: it counts pixel patterns, not meanings. Have them compare their own labeling process to the computer’s, highlighting the difference between human intuition and pattern matching.

Common Misconception

During Challenge: Data Quality Impact, watch for students assuming more data always improves results.

What to Teach Instead

Ask groups to test a clean dataset, then a noisy one, and measure error rates. Have them present findings to the class, showing how duplicates or mislabels degrade performance.

Common Misconception

During Hands-on: Unsupervised Clustering, notice students thinking the algorithm needs no data to find patterns.

What to Teach Instead

Have students physically shuffle cards and observe how groupings emerge only after data is introduced. Ask them to explain why the algorithm needs unlabeled data to self-organize.

Assessment Ideas

Exit Ticket

After Demo: Supervised Image Classifier and Hands-on: Unsupervised Clustering, provide three scenarios: 1) Identifying spam emails (labeled), 2) Grouping news articles by topic (unlabeled), 3) Predicting house prices (labeled). Ask students to write which type of learning would be best for each and explain why.

Quick Check

During Challenge: Data Quality Impact, present students with a simple dataset of fruits with colors and sizes. Ask them to imagine training a model to identify apples. What kind of data would they need? What would be the 'label' for supervised learning? How might they evaluate if the model is learning well?

Discussion Prompt

After Whole Class: Prediction Relay, pose the question: 'If you were building a system to recommend music, would you use supervised or unsupervised learning? What are the pros and cons of each for this specific task?' Encourage students to consider the type of data available and the desired outcome.

Extensions & Scaffolding

• Challenge: Ask students to design a new dataset with intentional biases and test its impact on a classifier during the Challenge: Data Quality Impact activity.
• Scaffolding: Provide pre-labeled examples for the Hands-on: Unsupervised Clustering activity if students struggle to start.
• Deeper exploration: Have students research how bias in training data leads to real-world discrimination, then discuss implications during the Whole Class: Prediction Relay.

Key Vocabulary

Training Data	A large set of examples, often with labels, used to teach a machine learning model to recognize patterns or make predictions.
Algorithm	A set of rules or instructions that a computer follows to perform a task, in machine learning, this is how the model learns from data.
Supervised Learning	A type of machine learning where the algorithm is trained on data that is already labeled with the correct answers, like pictures of cats labeled 'cat'.
Unsupervised Learning	A type of machine learning where the algorithm is given unlabeled data and must find patterns or structures on its own, such as grouping similar customers.
Model Parameters	The internal variables within a machine learning model that are adjusted during the training process to improve its performance.