Broadly speaking, a neural network is a type of machine learning framework that is built for pattern-matching. Neural networks are often used to classify input data, such as images or text, based on the extensional definitions of the type of object they are identifying. A classifier network, for example, might be given images as input and labels as output, and then use this to determine an internal function that will map the Inputs to the Outputs:
The first black box operation in the preceding diagram indicates that the network is being trained and is updating its approximation of the input and output function based on incoming data. The second box represents the testing process, where the network uses its internal function to make predictions on new incoming data.
Think of the way a machine learning algorithm, such as a decision tree classifier...