Skip to content

Exploring ML: testing accuracy

Lesson 4 of 8

CreateAI, MakeCode

11-14 yrs

This is the fourth of eight lessons and uses micro:bit CreateAI, a free tool for teaching about ML (machine learning) using movement data on the BBC micro:bit.

Students train and test a machine learning model for an AI exercise timer, then examine it for accuracy using a record card.

This lesson download includes required and optional materials, see lists below.

The Record card can be printed and copied for students to complete, or they can draw a similar table on paper to use instead.

Required materials for this lesson:

  • Lesson plan
  • Slides
  • Record card

Optional resources included in this lesson download:

  • Data and CreateAI worksheet

Key learning:

  • I can consider if the use of my data in CreateAI is safe and appropriate.
  • I can train and test my machine learning (ML) model to react to different kinds of movements.
  • I can examine the accuracy of my ML model using a record card.

AI literacy:

How machines learn

Human role in AI design

Testing ML models

Understanding AI

Data literacy:

Cleaning data

You may also be interested in

Two lessons introducing the concepts of AI and machine learning using micro:bit CreateAI.

Seven lessons introducing AI in general and unpacking what students already know about sorting data, before getting hands-on, creating, testing and improving their own machine learning models to be used as new inputs in code on the micro:bit.