A survey of the past, present and future of echo state networks

Along with the development of Machine Learning, statistic and Artificial Intelligence, people are exposed to myriad of big data. Meanwhile, accurate data analysis is difficult. Echo state network (ESN) algorithms are widely researched and applied in many fields. Owing to their potential for exact prediction and simple training process, scientists pay more attention to the research of ESN. In this paper, the representative research is carried out to sum up the research achievements on ESN, and the future development direction is discussed by pointing out the key technical challenges and we suggest several strategies for tackling the challenges.