Neural networks: foundations and mathematics

University-grade introductory path to neural networks, from the formal neuron definition all the way to regularisation and advanced optimisers. Language-agnostic, accessible to a motivated high-school reader.

1. 00
   Foreword

   Why neural networks, which problems they actually solve, and how to read this course.
   8 min
2. 01
   The artificial neuron

   From biological to mathematical, what really happens inside the elementary brick of a network.
   18 min
3. 02
   Essential linear algebra

   Vectors, dot products and matrices: exactly what you need to speak neural-network language.
   22 min
4. 03
   Activation functions

   Identity, sigmoid, ReLU, tanh: why they exist, what they give, and how to choose.
   16 min
5. 04
   The perceptron

   How Rosenblatt taught a machine to learn without a gradient (1958).
   35 min
6. 05
   From the neuron to the multilayer network

   Stacking neurons to solve XOR, then approximating almost any function.
   22 min
7. 06
   Forward pass and loss functions

   From input to prediction, then how to give it a score.
   20 min
8. 07
   Derivatives and the chain rule

   From the slope of a curve to the gradient: the tool that turns the cost landscape into a path of descent.
   18 min
9. 08
   Backpropagation

   A single backward pass that recovers the gradient of every weight: the chain rule, industrialised.
   22 min