Strengthening models through adding diverse data
Lesson 6 of 6
CreateAI, MakeCode
9-12 yrs
In this lesson students put their project on a micro:bit and allow other students to test it out. They identify how adding more data could strengthen the ML model by making it work better for more people.
Optionally, they can add more data from a more diverse range of people to their ML model, re-train and re-test it.
Key learning:
- I can evaluate an ML model and code running on a micro:bit using live data from different people.
- I can identify how to make an ML model more robust by adding more data from different people.
- I understand that ML models perform better if they have been trained on data from different groups of people.
AI literacy:
Human role in AI design
Impact of AI
Testing ML models
Data literacy:
Data bias
