Model-assisted deep learning of rare extreme events from partial observations
arXiv:2111.04857 · doi:10.1063/5.0077646
Abstract
To predict rare extreme events using deep neural networks, one encounters the so-called small data problem because even long-term observations often contain few extreme events. Here, we investigate a model-assisted framework where the training data is obtained from numerical simulations, as opposed to observations, with adequate samples from extreme events. However, to ensure the trained networks are applicable in practice, the training is not performed on the full simulation data; instead we only use a small subset of observable quantities which can be measured in practice. We investigate the feasibility of this model-assisted framework on three different dynamical systems (Rossler attractor, FitzHugh-Nagumo model, and a turbulent fluid flow) and three different deep neural network architectures (feedforward, long short-term memory, and reservoir computing). In each case, we study the prediction accuracy, robustness to noise, reproducibility under repeated training, and sensitivity to the type of input data. In particular, we find long short-term memory networks to be most robust to noise and to yield relatively accurate predictions, while requiring minimal fine-tuning of the hyperparameters.
arXiv commentAccepted for publication in Chaos: An Interdisciplinary Journal of Nonlinear Science
References in corpus (23)
- Fundamentals of Recurrent Neural Network (RNN) and Long Short-Term Memory (LSTM) Network
- Deep learning for time series classification: a review
- Machine Learning for Fluid Mechanics
- Machine learning accelerated computational fluid dynamics
- Using Machine Learning to Replicate Chaotic Attractors and Calculate Lyapunov Exponents from Data
- Data-assisted reduced-order modeling of extreme events in complex dynamical systems
- Analog forecasting of extreme-causing weather patterns using deep learning
- Robust flow field reconstruction from limited measurements via sparse representation
- Data-driven prediction of a multi-scale Lorenz 96 chaotic system using deep learning methods: Reservoir computing, ANN, and RNN-LSTM
- Extreme events in excitable systems and mechanisms of their generation
- Rogue Waves and Large Deviations in Deep Sea
- A variational approach to probing extreme events in turbulent dynamical systems
- Route to extreme events in excitable systems
- Sparse sensing and DMD based identification of flow regimes and bifurcations in complex flows
- An adjoint-based approach for finding invariant solutions of Navier-Stokes equations
- Dynamical indicators for the prediction of bursting phenomena in high-dimensional systems
- Reduced-order prediction of rogue waves in two-dimensional deep-water waves
- Using reservoir computer to predict and prevent extreme events
- Are extreme dissipation events predictable in turbulent fluid flows?
- Using Machine Learning to predict extreme events in the Hénon map
- Model-free prediction of emergence of extreme events in a parametrically driven nonlinear dynamical system by Deep Learning
- Investigating climate tipping points under various emission reduction and carbon capture scenarios with a stochastic climate model
- Data-driven prediction of multistable systems from sparse measurements