Combining machine learning and data assimilation to forecast dynamical systems from noisy partial observations

Abstract

We present a supervised learning method to learn the propagator map of a dynamical system from partial and noisy observations. In our computationally cheap and easy-to-implement framework a neural network consisting of random feature maps is trained sequentially by incoming observations within a data assimilation procedure. By employing Takens' embedding theorem, the network is trained on delay coordinates. We show that the combination of random feature maps and data assimilation, called RAFDA, outperforms standard random feature maps for which the dynamics is learned using batch data.

References in corpus (6)

• Hidden Physics Models: Machine Learning of Nonlinear Partial Differential Equations
• Next Generation Reservoir Computing
• Reservoir Computing as a Tool for Climate Predictability Studies
• Online learning of both state and dynamics using ensemble Kalman filters
• Leveraging Bayesian Analysis To Improve Accuracy of Approximate Models
• Forecasting Using Reservoir Computing: The Role of Generalized Synchronization