🕸️ Neural Network

5th Grade 🤖 Artificial Intelligence

A neural network is a kind of AI model built from layers of small calculating units connected to each other. The idea was inspired by nerve cells in the brain, but it works in a completely different way.

Free · US Letter · no sign-up
Google Classroom Save to Pinterest Email me

Each unit in the network receives numbers, multiplies them by “weights,” adds them up and passes the result on to the next layer. The first layer receives the input — for example, the pixels of a picture — and the last layer gives an answer, like “dog” or “cat.” During training, the weights are changed until the answers come out right.

In a network with many layers, each layer learns to spot something a little more complex: early layers find lines and edges, middle layers find shapes like a circle or an ear, and later layers find whole things like faces. Learning with many-layered networks is called “deep learning,” and it’s behind image recognition, speech recognition and language models.

A common mistake is thinking a neural network is an “artificial brain” that thinks like a person. Your brain is far more complex — it feels, experiences things and can learn from just a few examples. A neural network is a gigantic math formula with an enormous number of numbers, and even experts find it hard to explain exactly why it gave a particular answer.

For example

  • A neural network that reads handwritten ZIP codes on envelopes at the post office.
  • A network that takes a speech recording and outputs the words as text.
  • A network that helps doctors mark suspicious spots on X-rays.
  • A network that guesses the next word in a sentence — the basis of language models.

📌 Remember: Neural network = layers of small calculations, inspired by the brain but not a brain.

Quick quiz

  1. What is a neural network made of?

  2. What gets changed in the network during training?

  3. Which statement about neural networks is true?

🔎 Ask Thinkapedia another question