IA, apprentissage automatique et apprentissage profond

Think of AI as the whole family: any technique that makes a machine act smart. Everything you will learn lives somewhere inside it. 🤖

AI is the broad goal of getting computers to do tasks that normally need human smarts, like spotting a cat or translating a sentence.

Machine learning is one branch of AI. Instead of you writing every rule, the program learns patterns straight from examples in your data.

Show a model thousands of labeled emails and it learns to call new ones spam or not. That shift from rules to examples is the heart of ML.

Deep learning is a smaller circle inside machine learning. It uses layered neural networks to learn really complex patterns on its own.

The word deep just means many stacked layers. Each layer transforms the data a little, and together they capture rich, subtle structure.

Picture three nested circles: AI holds machine learning, which holds deep learning. Deep learning is a powerful slice of the whole field, not a rival.

Classic machine learning often needs you to hand it useful features, like word counts, before it can learn anything from raw data.

Deep learning skips much of that prep: it learns its own features directly from raw pixels or text, which is why it shines on messy data.

Deep models are hungry. They usually need lots of data and computing power, so a simpler ML model can win when examples are scarce.

Deep learning is a brilliant tool, not the only one. Knowing where it sits helps you choose the right approach for each problem you face.

Let us pin down how these three fields relate.

You placed the family tree: AI holds machine learning, which holds deep learning. Deep nets learn features from data, while simpler models still win when data is small. 🌳