Today, the neural computer attracts researchers' attention as the computer which aim at realizing excellent ability of information processing of a brain. The neural computer is based on neuron that is fundamental element of a brain and a sturucture of neural network that is linked with neurons togeter.
Last year, Ohno proposed neural network model of discarete time which learns chaotic attractor of difference equation, and he made sure of its effect.
In this paper, I propose the way in which neural network learns chaotic attractor with backpropagaion learning method, making sure of its effect. Further, in this paper I propose an algorithm in which neural network learns two chaotic attractors, making sure of its effect.
As an example of chaotic attractor, I used the 
map and map of folded pattern.
After learning, I ascertain that neural networks has acquired the nonliner dynamics of the map and map of folded pattern.