COURSE 05L1100% FREE
Verified 2026-08-10

Neural networks

A **neural network** is a stack of simple math units (neurons) that transform inputs into outputs. With enough layers and data, networks learn useful internal features—edges in images, patterns in sound, structure in text.

What it is

A neural network is a stack of simple math units (neurons) that transform inputs into outputs. With enough layers and data, networks learn useful internal features—edges in images, patterns in sound, structure in text.

Why it matters

Deep learning—today’s engine for vision, speech, and LLMs—is neural networks trained at scale.

How it works (plain)

Input numbers flow forward through layers. Each layer mixes inputs with learned weights and applies a nonlinearity. Training nudges weights so predictions match labels better (backpropagation—next chapters).

Everyday example

Image classifiers, speech-to-text engines, and transformers are all neural nets with different wiring diagrams.

Try it

Draw three boxes in a row labeled Input → Hidden → Output. That is the cartoon of a tiny network.

Myths

⚠️ Myth: Neural nets are digital brains.
✓ Reality: Biological metaphor inspired the name; the math is not a full brain model.
⚠️ Myth: Deeper always wins.
✓ Reality: Depth helps with the right architecture, data, and training tricks.

Sources