Chaotic Signal Denoising Based on Simplified Convolutional Denoising Auto-Encoder

Chaos is a ubiquitous phenomenon in nature, but the observed chaotic signals are often contaminated by noises. In this work, we consider chaotic signal denoising from the perspective of deep learning, and propose a chaotic signal denoising method referred to as Simplified Convolutional Denoising Auto-Encoder (SCDAE). The method consists of an encoder and a decoder with $13$ layers in total, and requires minimal preprocessing steps. Our simulation results show that the proposed method can achieve smaller root mean square errors and better proliferation exponents than conventional denoising techniques.