📄 To save as PDF: Press Ctrl+P (or ⌘P) → Destination: Save as PDF → Save
TeacherColab · Free Student Worksheet
Training vs Testing Data in Machine Learning
Student Name
Date
Learning Objective
Learn why AI models are evaluated on data they have never seen before — and why testing on training data is cheating.
Section 1 — What do you already know?
Before we begin, write down anything you already know about: Training vs Testing
Section 2 — Key Concepts
1. What is training data?
Answer:
Training data is the set of examples the machine learning model learns from. The model studies patterns in the training data during the training process.
2. What is testing data?
Answer:
Testing data is a separate set of examples the model has never seen. After training, the model is evaluated on the test set to measure how well it performs on new, unseen data.
3. Why can't we test on training data?
Answer:
Testing on training data is like giving students an exam using the exact questions from their textbook — they might get 100% by memorising, but that doesn't prove they truly understand. Similarly, an AI that tests on its training data may appear perfect but fail on real-world data.
Section 3 — Check Your Understanding
Answer the following questions in your own words.
1. Explain why data is split into training and testing sets
2. Define overfitting and explain why it is a problem
3. Apply the 80/20 train-test split rule
4. Understand model evaluation using a test set
Section 4 — Fill in the Blanks
Complete the sentences using the concepts you have learned.
1. Training vs Testing helps computers _________________ patterns from _________________.
2. The main difference between AI and a normal program is that AI _________________ from data rather than following fixed _________________.
3. One real-world example of Training vs Testing is _________________, which uses it to _________________.
4. The inputs used by an AI model are called _________________ and the output it predicts is called a _________________.
Section 5 — Scenario Thinking
🤔 Think about this:
A healthcare organisation wants to use Training vs Testing to assist doctors in diagnosing patients earlier. What data would the AI need? What challenges could arise? How can the team ensure the AI is fair, safe, and accurate?
Section 6 — What Did I Learn?
✅ Three key ideas I now understand:
❓ One question I still have:
⭐ Challenge Task
If you were designing an AI solution using Training vs Testing, what problem would you solve? Outline a 3-step plan: (1) Problem statement, (2) Data to collect, (3) How to verify accuracy.