Deep Learning

→ חזרה למאמר

Deep learning uses a neural network with many layers so it can learn harder patterns.

What it does

Deep learning uses a neural network with many layers so it can learn harder patterns.

How it works

  • Send data into the first layer.
  • Let each layer find a small pattern.
  • Combine many layers to make a final answer.

Uses and applications

  • A car system can notice lanes and signs in camera images.
  • A phone can understand speech commands.
  • A photo app can sort pictures by what is inside them.

In short

Deep learning is part of DQN, where a deep network looks at game screens.

Related methods

  • Neural Network: The basic building block deep learning stacks.
  • ResNet: Uses shortcuts so very deep networks can train.
  • Transformer: The architecture behind most current language models.

Back to Neural networks and deep learning · All topics

The interactive demonstration and the Python walkthrough for this topic need JavaScript.