The purpose of this perceptron is to perform a binary classification of the inputs using a supervised learning strategy.
The idea of a perceptron is to have a single neuron that receives inputs, processes them and generates an output.
Every input has its own importance (weight): this allows the neuron to calculate the weighted sum of the inputs.
The neuron will then apply a non-linear activation function to it.
This implementation is based on the first paragraph of ujjwalkarn's A Quick Introduction to Neural Networks.
Includes a visual representation of the perceptron's training.
Training data consists of the coordinates of points in a cartesian grid.