Open fare
From €1 you choose · up to €100
AI Line
Popular
AI From Zero: Build a Neural Network with NumPy
Understand how AI learns by building a neural network yourself, one line of NumPy at a time.
Rated 4.7 out of 5 from 3 reviews
- Beginner
- 18 lessons · 3h 5m of video
- 503 students
- English
- Updated Oct 2026
The route
Curriculum
5 modules · 18 lessons · 3h 5m of video
-
1
27m
Arrays and data
3 lessons
-
2
31m
A single neuron
3 lessons
-
3
39m
Layers and activations
4 lessons
-
4
46m
Loss and gradients
4 lessons
-
5
42m
Training on digits
4 lessons
Destination
What you will be able to do
- Represent data and weights with NumPy arrays
- Build layers and activation functions
- Compute loss and gradients by hand
- Train with gradient descent
- Recognise handwritten digits with your own network
Before you board
You will need
- Basic Python
- School-level maths; everything else is explained
About this course
The journey
Most AI courses start with a library that hides how learning works. This one starts with nothing but Python and NumPy and builds a neural network by hand: neurons, layers, loss, gradients and training.
By the end your own network recognises handwritten digits with over 95% accuracy, and you understand every line of the code that does it — which makes every framework afterwards much easier.
Frequently asked questions
Still unsure about something? Write to our contact page and a human answers, usually the same working day.
Beginner. No previous experience with this language is assumed.
No. You pay once and keep lifetime access, including every future update to this course.
Yes. The preview video at the top of this page shows real code from the course being written and run.
Checkout is handled on our provider's secure payment page. The moment your payment clears we email your personal access link and access code to the address you used at checkout, and the same link appears in your account library. There is nothing to install and nothing to wait for.
Email within 14 days of your purchase, quote your order number, and we refund the full amount to your original payment method. No form to fill in and no questions about how much of the course you watched.
What students say
Reviews are written by people who bought this course. We publish the critical ones too.
4.7
Rated 4.7 out of 5Course rating · 3 reviews
Rating distribution
- 5 stars 67%
- 4 stars 33%
- 3 stars 0%
- 2 stars 0%
- 1 star 0%
-
CJOct 2026
Clara Jensen
Data analyst
Rated 5.0 out of 5Backpropagation finally makes sense
Writing the gradients by hand with NumPy is what made it click. When I moved to PyTorch afterwards, I understood what every line was doing.
-
RMAug 2026
Ravi Menon
Software engineer
Rated 5.0 out of 596% on digits with my own network
Seeing my hand-written network recognise digits was a real moment. The maths is explained only as far as you need it.
-
SBJun 2026
Sara Beaumont
Biology researcher
Rated 4.0 out of 5Excellent, a bit fast at the end
The early chapters are wonderfully clear. The training chapter moves faster, so keep a notebook handy.
€21
EUR
One-time payment · lifetime access