Activity: "Which food do you like?"

Lego Spike Essential or Spike Prime Kit

gain a detailed understanding of artificial neural networks in a fun and gradual way

2-3 hours

from age 10

Students will build a very simple "smart throne" and solve a seriesof exercises: each time a new figurine sits on the throne, a neural network must be retrained so that the throne smiles or does not smile depending on the figurine's food preferences.

For example, as in "Joshua likes pizza," the throne must smile when Joshua is given a pizza: a connection must therefore be made between the input neuron "pizza sensor" and the output neuron "smile."

But how do you connect the network for "Hungry Larry," who wants pizza and vegetables? Or for "Alice does not mix," who wants either one or the other, but not both together?

Fortunately, rather than making connections "by hand" in the neural network, we can also let the AI'slearning ability find the right connections on its own!

Details of the activity can be found in the PDF document at the top right. It is currently only available in English.

First exercise: connect the "pizza sensor" input neuron to the "smile" output neuron to make the brick smile when pizza is present.

‍

Second exercise: adding a vegetable sensor and connecting the neurons for Isabelle, who wants vegetables

Seventh (and final) exercise: multi-layer network for "Alice does not mix," with the help of machine learning!

‍

Download

"Which food do you like" presentation article

Related courses

Your basket

Discover our
resources