🧠
课程 2 · 年级 8+
Neural networks: how the machine learns
📖 课程
──▶
🎯 测验——选择翻译
──▶
✅ 本课已完成
A neural network is inspired by the brain: millions of small "neurons" in layers. Each neuron is a simple formula: input × weight. Training means tuning billions of weights so the error shrinks. Show a picture, the net says "dog", correct is "cat" — weights nudge toward "cat". Repeat a million times — and the net "knows".
- ●Input layer: pixels, words, numbers
- ●Hidden layers: find features — edges, shapes, meanings
- ●Output: the answer — "cat", translation, next move
⚡ Training a large model costs millions of euros in electricity — so only big companies and labs build them.