Year 9 - Computing
Documenting and evaluating a machine learning model
Machine learning using the micro:bit
Study Hub
Study Guides
This pupil-friendly study guide includes a unit summary, clear notes, common mistakes, and quick questions with answers.
Unit Summary
This unit is designed to develop pupils programming skills in a new engaging context: physical computing, using the BBC micro:bit. They will get acquainted with the host of components built into the micro:bit, and write simple programs that use these components to interact with the physical world.
Lesson Summary
You will learn to create a model card to document a machine learning model.
Key Notes
- It is important to evaluate a machine learning model's accuracy. A confusion matrix can be used to visualise this.
- Reporting a model's accuracy improves accountability so we can be more open about mistakes and issues.
- A model card provides more information, improving accountability and demonstrates responsible practice.
Vocabulary To Learn
- accuracy: how many times a model makes a correct prediction
- accountability: taking responsibility for how a model behaves and being honest about its performance, limitations and potential impact on others
- model card: a short report on how a machine learning model works, how well the model performs and what people should know before using the model
Common Mistakes To Avoid
- Users of a machine learning application do not need to be aware of how accurate it is.
3 Quick Questions (With Answers)
1. Describe the system or process from this lesson in clear steps.
It is important to evaluate a machine learning model's accuracy. A confusion matrix can be used to visualise this. Reporting a model's accuracy improves accountability so we can be more open about mistakes and issues.
2. Define this computing term and give one practical example. 'accuracy'
how many times a model makes a correct prediction. Add one real device or system example to prove understanding.
3. Correct this common computing misconception.
Mistake: Users of a machine learning application do not need to be aware of how accurate it is. Correction: Model cards are good practice when developing machine learning applications. A model card is a short report on how a machine learning model works, how well the model performs and what people should know before using the model.
More Lessons In This Unit
Browse all guides in the Year 9 Computing guide library.