One Line Is Not Enough
Discover how artificial brains learn patterns.
What you'll learn in this level
- What weights between neurons do
- Why the XOR problem can't be solved with one line
- What a hidden layer is actually good for
Neurons as a Team
Neural networks are inspired by how your brain works! Just like your brain has billions of connected neurons, a neural network has layers of artificial neurons.
These neurons work together, passing signals and learning patterns. By adjusting the connection strength (weights), the network can learn to solve complex problems.
How It Works
- 1
Layers of Neurons
The input layer takes in the data. Hidden layers work with it. The output layer gives the answer.
- 2
Learning by Weights
The network changes how strong its links are. That is how it gets better.
- 3
Fire or Not?
Every neuron adds up its inputs. Then it decides whether to fire. That rule is called an activation function.
- 4
Real World
This is how image recognition, translators and self-driving cars work.
Ready? Now you train your own network.
Guess first, then experiment
What is inside a single artificial neuron?