📄️ Table of contents
Free discovery course on Deep Learning in 6 lessons. What a neuron computes, why depth changes everything, how CNNs see images and transformers read text, what training a network really costs, and why you should almost never start from scratch.
📄️ 1. What a neuron computes
Lesson 1 of the free Deep Learning course: what a single artificial neuron does, why weights and biases are the learned parameters, what activation functions are for, and how a layer becomes a matrix multiplication.
📄️ 2. Why depth changes everything
Lesson 2 of the free Deep Learning course: how a hierarchy of representations emerges layer by layer, why this removes the need for feature engineering, what residual connections fixed, and the price depth charges.
📄️ 3. CNNs: how machines see
Lesson 3 of the free Deep Learning course: what convolution does, why weight sharing makes image recognition tractable, what pooling is for, and why a CNN recognises an object regardless of where it appears.
📄️ 4. Sequences and transformers
Lesson 4 of the free Deep Learning course: why recurrent networks struggled with long sequences, what LSTM fixed, what attention does, and the two properties that made transformers take over the whole field.
📄️ 5. What it really costs
Lesson 5 of the free Deep Learning course: the real data, compute and time requirements, why training from scratch is almost never the right choice, how transfer learning works, and when deep learning is the wrong tool.
📄️ 6. Recap and FAQ
Lesson 6 of the free Deep Learning course: a one-page synthesis, a table matching data types to architectures, a glossary of the terms you will meet, and answers to the 12 questions asked most.
📄️ 7. Quiz and attestation
Test your understanding of deep learning with a free 5-question quiz, corrected and explained: what a neuron computes, why activations matter, convolution and weight sharing, attention, and transfer learning.