Inspired by the brain, built from simple maths.
Direct answer
A neural network is made of layers of simple units. Each unit takes inputs, multiplies them by weights, adds them up and passes the result through a function. By adjusting the weights to reduce errors on training data, the network learns patterns such as recognising images or predicting text.
The pieces
- Input layer: receives the data.
- Hidden layers: transform it step by step.
- Output layer: gives the prediction.
- Weights: numbers that are adjusted during training.
How it learns
- Make a prediction.
- Measure the error with a loss function.
- Adjust weights using gradient descent.
- Repeat over many examples.
Where they are used
Image recognition, speech, translation and language models.
Where Teyro fits
Teyro's Machine Learning for Beginners course turns this into short lessons you can play in a few minutes a day. It has 200 lessons, and you can start free.
The bottom line
Many small calculations, tuned together, can learn very complex patterns.
