Phần mềm-Ứng dụng

An Artificial Neural Network (ANN) [3] simulates a biological neural network. An ANN is a block of interconnected process units in which the links between the neurals determine the function of the network. The processor unit is also called a neural or node, doing a very simple job: it receives input signals from front units or an external source and uses them to calculate the signal. This signal will be spread to other units. In an ANN, there are three types of units: input units, which receive signals from outside,  output units that send data to outside, and hidden units having their inputs and output in the network. Each unit can have one or more inputs, but only one output. An input to a unit may be data from outside the network, or an output from the same or from another unit.

To use an ANN, we need two data sets: a training data set which is a set of examples used for learning, that is to fit the parameters of the network, and a  test data set used only to assess the performance [generalization] of a fully specified network.