Machine Learning for Kids with KodeKiddo

The best way to teach kids about machine learning is through hands-on
experiences, and by making it fun using examples close to their interests.
What is machine learning? We probably use machine-learning systems every day without even knowing it. The technology gives us spam filters, our Facebook News Feeds, digital assistants, search engines, online shopping recommendations, and more. Machine learning is a subset of artificial intelligence which lets computer systems to perform a specific task without using explicit instructions, relying on patterns and inference instead.

While computers historically relied on people to tell them what to do and how to react, developers are creating machines that are able to learn. This means the robots can make their own decisions, giving them artificial intelligence.

To introduce this concept to children, during the Holiday Program and extra
curricular program, KodeKiddo held a course with the topic of “Machine Learning for Kids” for middle school and high school kiddos. Each project is done in Scratch, a widely used educational coding platform, and using IBM Cloud services to access Watson Assistant, IBM’s Machine Learning system. This course goes through step-by-step guide to create several programs using concepts of machine learning. Samples of the projects are such a program which makes a character that smiles if you say nice things to it and cries if you insult it, recognizes different hand positions when playing rock, paper, scissors, or makes recommendations based on certain data entered.

In this course, the kiddos learned the basic concepts of machine learning and implemented them in their coding to understand how machine learning can make their programs work better. In the first steps, they gathered data samples to train the program. After that, they created User Interface (UI) using Scratch and incorporate the additional command blocks in Scratch for the implementation of machine learning tasks, including using APIs from IBM Watson

In simple terms, here are the three steps: 

  1. Collect examples of things you want to be able to recognise
  2. Use the examples to train a computer to be able to recognise them
  3. Make a game in Scratch that uses the computer’s ability to recognise them

At the end of the program, the kiddos worked in a team to create one personal project using the concepts they learned and present it to the parents. See some examples below.

These kiddos are part of KodeKiddo extracurricular coding program at @LEC Laurensia Enrichment Center in Santa Laurensia school, Alam Sutera. They learned how to train simple machine-learning models and use them to play games and complete interactive projects in Scratch.

With the emergence of Artificial Intelligence technology around us, even kiddos should learn this machine learning concept earlier! 

Videos from the LEC Programs

Videos from the Holiday Program – Machine Learning Theme


Learn the Fundamental of Machine Learning. Download this Machine Learning for Dummies eBook from IBM.