In my PhD research, I implemented many echo state networks (ESNs) and tested them for a wide range of time series forecasting tasks. But I always recieved this questions that what are the ESNs and how do they fit into the deep learning context. This has always been surprising to me because this type of networks are very successful and efficient in predicting nonlinear and complex time series, and can be applied to solve a wide range of real-world problems.

This post will be updated after adding the ESN package in Downloads section.