Machine Learning for School Students, Netherlands

In short

Modern Age Coders teaches machine learning live online to school pupils in the Netherlands aged ten to eighteen. Younger pupils write a first program that learns from examples; older pupils train models in Python on real data, test them against a simple baseline, count what the model gets right and wrong in each group, and write up what it can and cannot do. Pupils are placed by ability, in groups of five to ten or one to one, alongside a Dutch or international school. The first lesson is free; after that a group place is USD 100 a month and one-to-one lessons USD 150.

Start here

Three ways into machine learning for school pupils

One for a younger pupil ready to type, one for a teenager ready for real models, and one for a teenager who wants to explore machine learning visually before coding it. Each card opens its own syllabus and timetable.

ML / 01

Python and AI for Kids

For pupils of nine to twelve: typed Python and a first program that learns a rule from examples, then fails on one the pupil chose on purpose.

Open the syllabus →

ML / 02

AI and Machine Learning for Teens

Statistics, classical models, neural networks, images and text, every model tested against a baseline and explained.

Open the syllabus →

Orange Data Mining Course for Teens course thumbnail

ML / 03

Orange Data Mining for Teens

Machine learning through visual workflows, a way to understand training and testing before writing the code.

Open the syllabus →

AI in Dutch schools

A national lab for AI in schools, which starts from the schools' own questions

NOLAI, the Nationaal Onderwijslab AI, is based at Radboud University in Nijmegen. The figures below come from its own pages and from the National Growth Fund, which finances it.

NOLAI in its own words and the fund's
Item	Detail	Source
What it is	The National Education Lab AI for elementary, secondary and special needs education in the Netherlands	Radboud University
How a project starts	With a question asked by a school, developed into a proposal with business and scientific partners	Radboud University
Funding	The National Growth Fund, and the European Union's NextGenerationEU	Radboud University
Amount conditionally awarded	51.1 million euros, first round	National Growth Fund
Period	From 2022 to 2035	National Growth Fund
Goals	77 co-creation projects, 15 validated intelligent educational technologies, at least 9 AI products shown in practice	National Growth Fund
Sharing	Proceeds shared open source as much as possible from 2026	Radboud University

The lab brings together pedagogy, ethics, data and AI technology, and describes co-operation between education, business and science as its key principle. It works on dozens of projects with schools across the country, and the fund page adds a second aim: making the pedagogical, social and societal consequences of digital innovation in education visible.

For a parent, the point is simple. AI is becoming part of how Dutch pupils are taught and assessed, and the country is investing seriously in doing that responsibly.

What that work does not do, and is not meant to do, is teach pupils to build AI themselves. A pupil who uses an adaptive practice program learns the subject it teaches; a pupil who builds a small model learns how such programs decide, why they can be wrong and what questions to ask of them.

That is the gap our classes fill. Informatica, the school subject closest to it, is optional and offered at a minority of schools; the national academy page gives the figures.

Sources: Radboud University, NOLAI; Nationaal Groeifonds, Nationaal Onderwijslab AI. Read 15 September 2026. We have no connection with NOLAI, Radboud University or the fund.

What pupils build

What a school pupil can build, and the test that separates a real model from a lucky one

Machine learning at school age is not a watered-down version of the adult subject. It is the same core ideas, learned on smaller problems, with one habit above all: never trust a model until it has beaten a simple guess.

Ages 10 to 12: a program that learns

A pupil gives a program labelled examples, such as short and long words or cats and dogs described by a few numbers, and lets it guess the label of a new one by finding the most similar example. Then the pupil chooses a case that fools it. The lesson, that a model only knows what its examples showed it, lasts for years.

Ages 13 to 15: train, then test on new data

The pupil splits real data into a part the model learns from and a part it never sees, and measures the model only on the unseen part. They also write down the score of the simplest possible guess. A model that cannot beat that guess has learned nothing useful, however clever it looks.

Ages 16 to 18: count what it misses

Older pupils look past a single score. They count, for each kind of case, how many the model caught and how many it wrongly flagged, and they write a short account of what the model is for, what data it learned from and where it fails. That account is often the strongest part of a profile project.

An invented example: 200 plant photos, 180 healthy and 20 diseased
Model	Diseased plants caught	Healthy plants wrongly flagged	Share of photos labelled correctly
Always says healthy	0 of 20	0 of 180	180 of 200, 90 percent
A trained model	15 of 20	12 of 180	183 of 200, 91.5 percent

The first row is the trap. A model that never finds a diseased plant still scores 90 percent, because nine photos in ten show healthy plants. Anyone who reported only that single figure would be praising a model that does nothing.

The second row is barely higher overall, 91.5 percent against 90, yet it is far more useful: it finds three quarters of the diseased plants, at the cost of wrongly flagging a few healthy ones. The arithmetic is 15 plus 168 correct healthy labels, which makes 183 of 200.

Seeing that once, on numbers a pupil worked out themselves, is worth more than any definition. From then on the pupil asks of every model: compared with what, and for which cases?

Our pupils also meet the same idea on real Dutch data; the page on building real AI projects in the Netherlands shows the public sources they use.

The plant example is invented for teaching and describes no real dataset. The age bands are our teaching plan, not a national standard.

Beside school

How machine learning fits beside a Dutch school career

Pupils learn this alongside their schoolwork, not instead of it. The timing below is what usually works.

For a pupil at the end of the basisschool, the first learning program is a natural next step after Scratch or first Python. The kids page covers that stage.

Between the brugklas and the third year, pupils build the Python and data skills a real model needs. The profielkeuze is a good moment to try a first trained model, because it shows quickly whether data and technology appeal.

In the upper years of havo and vwo, a model with an honest write-up makes a strong profile project or portfolio piece. The teens page covers the wider teenage route.

For pupils thinking about a university AI degree, the Groningen academy page sets out what a first year asks.

The rungs for pupils

From a program that guesses to a model that is tested fairly

Each rung prepares the next. A pupil who has never split data cannot test a model honestly, and one who has never written Python cannot see inside the model at all.

The machine learning route for school pupils
Rung	Usually	What the pupil can do
1. Python by hand	Ages 10 to 12	Write small programs, including one that learns from examples
2. Data	Ages 12 to 14	Clean, plot and split real data into learning and testing parts
3. First models	Ages 13 to 16	Train a model and compare it with the simplest guess
4. Deeper models	Ages 15 to 18	Neural networks, images and text, each tested fairly
5. Write-up	Throughout	A public repository with the model, its data and its limits

Why the simple guess comes first

Pupils love a high score. Teaching them to write down the score of the dullest possible guess before training anything protects them from the most common mistake in machine learning, at any age.

The complete order of subjects is on the coding roadmap.

Around toetsweken and exams

Most pupils take two lessons a week and pause during school test weeks. In the eindexamen year, many keep lessons light or stop until the exams are over.

Pupils at international schools follow the same rungs, placed by ability rather than by grade.

The catalogue

Eight live courses for pupils learning machine learning

Grouped by stage. Every card leads to the full course with its schedule and fee, and the free lesson settles where a pupil starts.

I

Ages 9 to 12

First code and a first learning program

ML / YOUNG / 01

Python and AI for Kids

Typed Python and a small program that learns from examples, for pupils in the last years of primary school.

Open the syllabus

ML / YOUNG / 02

AI Literacy for Kids

How AI systems learn and why they get things wrong, the understanding a young pupil needs before building one.

Open the syllabus

II

Ages 12 to 15

Python, data and a first fair test

ML / MID / 01

Python for Teens

The Python foundation every machine learning project depends on, built over two years.

Open the syllabus

ML / MID / 02

Data Science for Teens

Real datasets, cleaning, plotting and the split between learning and testing data.

Open the syllabus

ML / MID / 03

Orange Data Mining for Teens

Training and testing models through visual workflows, before or alongside code.

Open the syllabus

III

Ages 15 to 18

Real models, honestly tested

ML / UPPER / 01

AI and Machine Learning for Teens

Classical models to neural networks, each compared with a baseline and written up.

Open the syllabus

ML / UPPER / 02

AI Tools for Teens

Using AI tools well in projects and study, with the judgement a model-builder brings.

Open the syllabus

ML / UPPER / 03

Git and GitHub for Teens

A public home for every model and its write-up, from the first commit onwards.

Open the syllabus

How the classes run

After school, at home, with a teacher watching the model train

Our teachers work from India, which never switches to summer time, while Dutch clocks do. The gap is therefore three and a half hours in summer and four and a half in winter, which puts pupils' lessons comfortably in the late afternoon or at the weekend. The regular time is chosen with the family during the free lesson.

After school

Late afternoon, Dutch time. The usual choice for pupils of every age.

Early evening

By arrangement, for older pupils with sport or a part-time job in the afternoon.

Weekend

Saturday or Sunday, for longer sessions on a project.

A first lesson with a real model

Even in the free lesson a pupil makes something that learns. The teacher watches how they reason and chooses the right starting rung.

Groups by level

Five to ten pupils from several countries who can already do the same things, so nobody is rushed or held back.

One to one for projects

For a profile project, a competition entry or a pupil who prefers a teacher to themselves.

Parents informed

We speak with a parent before anything is booked and keep them in the loop, as Dutch rules on consent for under-sixteens expect.

English, with Dutch school words

Lessons are in English; brugklas, havo, vwo and profielkeuze are used as families use them.

Models kept and explained

Every model goes into the pupil's own GitHub repository with a short note on its data, its test and its limits.

Student work

Work our students have published

Four projects our students built and left online. More, including school-age work, are on the student labs page.

AI and ML

NutriLife

An AI nutrition coach that reads what you eat and works you toward a target.

by Bhavya · Open it

AI and ML

Misti

A chatbot that answers mathematics and programming questions, built and deployed by a student.

by Harshit · Open it

AI and ML

GuardianX

An assistant that helps a young person recognise unsafe situations online.

by Vivaan · Open it

Web app

SkyCast

A weather forecasting site with live conditions for any location.

by Krish

Fees

Fees for school pupils

Monthly, in US dollars, the same as for every country outside India. No separate Dutch rate, no joining fee, no annual contract.

Free first class

USD 0

no card required

A real lesson with a real model
Places the pupil at the right rung
No card or bank details

Book it

Group batch

USD 100

a month, billed in US dollars

Five to ten pupils at one level
The same live teacher every week
Models reviewed every week
A certificate at the end

Start here

One to one

USD 150

a month, billed in US dollars

A teacher for one pupil
Pace set by the pupil and the project
Ideal for profile projects

Enquire

What families say

Rated 4.9 across 547 Google reviews

Quoted from Google as parents and students wrote them. We do not edit or pay for reviews.

★★★★★

"The one step solution for my son. Modern Age Coders make learning coding so simple that kids love it. The teachers explain complex concepts clearly with practical exercises and interactive content."

Ria Mukherjee

Parent

★★★★★

"Modern Age Coders has been a game-changer for me. I struggled to grasp IT concepts and coding before joining, but their classes transformed everything. I can now confidently write complex programs with ease."

Samriddha Mondal

Student

★★★★★

"One of the most wonderful education centres out there. Education is not limited to school syllabus but focuses on skill development."

Vansh Agarwal

Student

★★★★★

"My child Dhairya is really enjoying the Modern Age Coders classes. This is his first online class and he eagerly looks forward to it. I can already see his improvement, and the teachers are very cooperative."

Sonam Oswal

Parent of Dhairya

★★★★★

"Modern Age Coders have wonderful teachers who teach in a clear, easy and practical way. The teacher boosts students' confidence and inspires them to learn without hesitation."

Sonu Goyal

Parent

★★★★★

"I highly recommend this computer coding class! The teachers are incredibly knowledgeable and passionate about coding."

Ritu Kedia

Parent

Questions from parents and pupils

What Dutch families ask about machine learning at school age

Is machine learning too advanced for a school pupil?

Not in the form we teach it. The core ideas, learning from examples, testing on unseen data and comparing with a simple guess, can be learned from about ten. The mathematics deepens as the pupil does.

What is NOLAI, and is this connected to it?

NOLAI is the national education lab for AI at Radboud University, working with schools on AI used in education. We have no connection with it. It studies AI that pupils use; our classes teach pupils to build and test models themselves.

Why can a 90 percent accurate model be useless?

Because the simplest guess may already score that. If nine cases in ten belong to one group, a model that always picks that group is 90 percent accurate and never finds the rest. Our pupils always compare a model with that simple guess.

Does my child need informatica at school?

No. Many Dutch schools do not offer it, and our classes start from the beginning with Python. For a pupil who does take informatica, the classes go further into machine learning.

Can this become a profile project?

Yes. A trained model with a clear write-up of its data, its test and its limits makes a strong profielwerkstuk. We teach the skills; the project and its conclusions are the pupil's own.

What does a pupil need at home?

A laptop or desktop and a steady internet connection. No phone and no special software are needed at the start.

Are the lessons in English?

Yes. Python and its libraries are in English anyway, and teachers keep explanations clear for pupils at Dutch schools.

What does it cost?

A group place is USD 100 a month and one-to-one lessons are USD 150 a month, billed in US dollars, with no joining fee and no yearly contract. The first lesson is free.

When are lessons held?

After school or at the weekend, Dutch time. The teacher is three and a half hours ahead of Dutch summer time and four and a half ahead of winter time.

What happens after we send the form?

We call a parent at a Dutch hour to arrange the free lesson. There is nothing to pay unless the family decides to continue afterwards.

Elsewhere on this site

More for pupils and parents in the Netherlands

Pages for the stages before and after, and for the projects pupils build.

Coding classes for kids

The stage before, for children aged 6 to 12.

Coding classes for teens

The wider teenage route beyond machine learning.

Build real AI projects

Projects on real Dutch open data.

AI and Python Academy, Groningen

What a university AI first year asks of a pupil.

AI and Python Academy, Netherlands

The national ladder and the informatica figures.

AI and ML course for teens

The global teenage machine learning course page.

Start here

Book a free machine learning lesson

Leave a parent's number and we will call at a Dutch hour. In the free lesson your child builds something that learns, with a real teacher, and by the end we know where they should start.

Prefer to read first? See the AI and ML course for teens, how we teach, and the coding roadmap.

WhatsApp us · +91 91233 66161 · contact@modernagecoders.com

A WhatsApp message from a Dutch phone is free and gets the fastest reply. The number is Indian; we have no office or premises in the Netherlands.

Three ways into machine learning for school pupils

Python and AI for Kids

AI and Machine Learning for Teens

Orange Data Mining for Teens

A national lab for AI in schools, which starts from the schools' own questions

What a school pupil can build, and the test that separates a real model from a lucky one

Ages 10 to 12: a program that learns

Ages 13 to 15: train, then test on new data

Ages 16 to 18: count what it misses

How machine learning fits beside a Dutch school career

From a program that guesses to a model that is tested fairly

Why the simple guess comes first

Around toetsweken and exams

Eight live courses for pupils learning machine learning

Ages 9 to 12

Python and AI for Kids

AI Literacy for Kids

Ages 12 to 15

Python for Teens

Data Science for Teens

Orange Data Mining for Teens

Ages 15 to 18

AI and Machine Learning for Teens

AI Tools for Teens

Git and GitHub for Teens

After school, at home, with a teacher watching the model train

A first lesson with a real model

Groups by level

One to one for projects

Parents informed

English, with Dutch school words

Models kept and explained

Work our students have published

NutriLife

Misti

GuardianX

SkyCast

Fees for school pupils

Rated 4.9 across 547 Google reviews

What Dutch families ask about machine learning at school age

Is machine learning too advanced for a school pupil?

What is NOLAI, and is this connected to it?

Why can a 90 percent accurate model be useless?

Does my child need informatica at school?

Can this become a profile project?

What does a pupil need at home?

Are the lessons in English?

What does it cost?

When are lessons held?

What happens after we send the form?

More for pupils and parents in the Netherlands

Coding classes for kids

Coding classes for teens

Build real AI projects

AI and Python Academy, Groningen

AI and Python Academy, Netherlands

AI and ML course for teens

Book a free machine learning lesson

Keep exploring Modern Age Coders

Related courses and guides

Learn more

Free resources

From the blog

Start here