Mod-DeepESN: Modular Deep Echo State Network
This work addresses a domain-specific problem for researchers and practitioners in time series prediction, offering an incremental improvement over existing methods.
The authors tackled the problem of echo state networks underperforming on complex multi-scale temporal tasks by proposing a modular deep echo state network architecture infused with intrinsic plasticity, which outperforms state-of-the-art methods for time series prediction.
Neuro-inspired recurrent neural network algorithms, such as echo state networks, are computationally lightweight and thereby map well onto untethered devices. The baseline echo state network algorithms are shown to be efficient in solving small-scale spatio-temporal problems. However, they underperform for complex tasks that are characterized by multi-scale structures. In this research, an intrinsic plasticity-infused modular deep echo state network architecture is proposed to solve complex and multiple timescale temporal tasks. It outperforms state-of-the-art for time series prediction tasks.