Radar plots discrimination with neural networks

This paper investigates and compares two statistical methods with supervised learning to compute the confidence level from the available attributes of a given plot. The first is a classical Bayes rule; the second is a neural network trained via the backpropagation algorithm.<
>