Vanya Dimitrova Markova
Ventseslav Kirilov Shopov
Institute of Robotics, Bulgarian Academy of Sciences, Plovdiv
Bulgaria
e-mail: markovavanya@yahoo.com
Abstract:
In this study, we use deep recurrent neural networks to simulate the behaviour of dynamic chaotic attractors and to predict the next state of the system. We compare three dynamic attractors: Ueda, Pickover, and Burke-Shaw. Also, we study the ability to predict such time series simple Recurrent Neural Networks with Long Short Term Memory and Gated Recurrent Unit.
Key words:
recurrent neural networks
deep learning
chaotic attractors
Section:
Topics: