Meta-SysId: A Meta-Learning Approach for Simultaneous Identification and Prediction
This addresses the challenge of efficiently modeling related systems in domains like control and prediction, though it appears incremental as it builds on existing meta-learning methods.
The paper tackles the problem of modeling multiple systems with shared underlying laws but different contexts by proposing Meta-SysId, a meta-learning approach that separates shared parameters from context variables. It shows competitive or superior performance compared to baselines in polynomial regression, time-series prediction, model-based control, and traffic prediction tasks.
In this paper, we propose Meta-SysId, a meta-learning approach to model sets of systems that have behavior governed by common but unknown laws and that differentiate themselves by their context. Inspired by classical modeling-and-identification approaches, Meta-SysId learns to represent the common law through shared parameters and relies on online optimization to compute system-specific context. Compared to optimization-based meta-learning methods, the separation between class parameters and context variables reduces the computational burden while allowing batch computations and a simple training scheme. We test Meta-SysId on polynomial regression, time-series prediction, model-based control, and real-world traffic prediction domains, empirically finding it outperforms or is competitive with meta-learning baselines.