Is it possible to generate "chaotic" graph using machine learning?

Hey guys. I made a numerical simulation of the “chua circuit”, thus generating a “chaotic” graph.

In the next step, I would like to use machine learning to map the behavior of curves.

In the circuit simulation, 3 variables were generated, I already did the heavy part of creating Variavel_ENTRADA and Variavel_SAIDA.

Studying I discovered that it is possible to do the project with RECURRING NEURAL NETWORKS.

I saw that TensorFlow facilitates the process, someone could give me a “light” with an article that is close so I can move forward with the project.