Brain Builder
Understand how neurons connect and learn.
What you'll learn in this level
- How a net is built from neurons and layers
- What the input layer, hidden layers and output layer do
- What a weight is and why it changes while the net learns
What Is a Neural Net?
A neural network is a computer program built a bit like a brain. It is made of many small parts. They are called neurons.
Each neuron passes its message on to the next one. That way information travels through the whole net.
How a Net Works
- 1
Input Layer
This is where data comes into the net. It works like your eyes. They see something and pass it on.
- 2
Hidden Layers
These middle layers work on the data. Your brain does this too. It thinks about what you see.
- 3
Output Layer
Here the net gives its answer. It is like the moment you know what you are looking at.
- 4
Weights
Every connection carries a number. The number says how important the connection is. Learning means the net changes exactly these numbers.
Between two neurons sits a connection, and each one carries a number: the weight. It says how important that connection is. When the net learns, it changes these numbers.
Guess first, then experiment
You build a net with several layers. Which layer "sees" the data first?