Sequential Learning from Noisy Data: Data-Assimilation Meets Echo-State Network

Debdipta Goswami

This paper explores the problem of training a recurrent neural network from noisy data. While neural network based dynamic predictors perform well with noise-free training data, pr…