Meta-Learning Dynamics Forecasting Using Task Inference
arXiv:2102.10271
Abstract
Current deep learning models for dynamics forecasting struggle with generalization. They can only forecast in a specific domain and fail when applied to systems with different parameters, external forces, or boundary conditions. We propose a model-based meta-learning method called DyAd which can generalize across heterogeneous domains by partitioning them into different tasks. DyAd has two parts: an encoder which infers the time-invariant hidden features of the task with weak supervision, and a forecaster which learns the shared dynamics of the entire domain. The encoder adapts and controls the forecaster during inference using adaptive instance normalization and adaptive padding. Theoretically, we prove that the generalization error of such procedure is related to the task relatedness in the source domain, as well as the domain differences between source and target. Experimentally, we demonstrate that our model outperforms state-of-the-art approaches on both turbulent flow and real-world ocean data forecasting tasks.
References in corpus (20)
- Model-Agnostic Meta-Learning for Fast Adaptation of Deep Networks
- Prototypical Networks for Few-shot Learning
- Physics Informed Deep Learning (Part I): Data-driven Solutions of Nonlinear Partial Differential Equations
- Challenging Common Assumptions in the Unsupervised Learning of Disentangled Representations
- A Learned Representation For Artistic Style
- Decomposing Motion and Content for Natural Video Sequence Prediction
- A Review on Deep Learning Techniques for Video Prediction
- Accelerating Eulerian Fluid Simulation With Convolutional Networks
- Meta-Reinforcement Learning of Structured Exploration Strategies
- Recasting Gradient-Based Meta-Learning as Hierarchical Bayes
- Visual Dynamics: Probabilistic Future Frame Synthesis via Cross Convolutional Networks
- Hierarchically Structured Meta-learning
- Deep unsupervised learning of turbulence for inflow generation at various Reynolds numbers
- Deep Dynamical Modeling and Control of Unsteady Fluid Flows
- The Benefit of Multitask Representation Learning
- Incorporating Symmetry into Deep Dynamics Models for Improved Generalization
- Learning Dynamical Systems from Partial Observations
- Modular meta-learning
- Physics-aware Spatiotemporal Modules with Auxiliary Tasks for Meta-Learning
- Learning Disentangled Representations of Video with Missing Data