Neural network
What is Neural network?
A structure that learns by passing data through layers of simple connected units, loosely modelled on how brain cells signal each other. Each layer picks out patterns and hands them to the next. Stack enough layers and you get deep learning, which is the basis of almost all modern AI.
Why it matters
A neural network is the structure that lets a machine learn patterns instead of following written rules. Data passes through layers of simple connected units, loosely modelled on how brain cells signal each other, and each layer picks out patterns and passes them to the next. It matters because this design is the foundation almost all modern AI is built on. Stack enough layers and you get deep learning, and from there the language models and image tools businesses now use. You do not need to build one to use AI, but the term explains what is doing the learning underneath.
In practice
You rarely touch a neural network directly. It sits inside the products you buy: the model that reads a scanned document, the one that recommends the next film, the one behind a chatbot. When a vendor talks about model “size” or “parameters”, they are describing the scale of the network, the number of connections it tunes during training. More connections can mean more capability, and also more cost to run.
