Embed and Emulate: Learning to estimate parameters of dynamical systems with uncertainty quantification

Ruoxi Jiang, Rebecca Willett

Keywords: Learned emulators, contrastive learning, parameter estimation, inverse problems, climate forecasting

Abstract: This paper explores learning emulators for parameter estimation with uncertainty estimation of high-dimensional dynamical systems. We assume access to a computationally complex simulator that inputs a candidate parameter and outputs a corresponding multi-channel time series. Our task is to accurately estimate a range of likely values of the underlying parameters. Standard iterative approaches necessitate running the simulator many times, which is computationally prohibitive. This paper describes a novel framework for learning feature embeddings of observed dynamics jointly with an emulator that can replace high-cost simulators. Leveraging a contrastive learning approach, our method exploits intrinsic data properties within and across parameter and trajectory domains. On a coupled 396-dimensional multiscale Lorenz 96 system, our method significantly outperforms a typical parameter estimation method based on predefined metrics and a classical numerical simulator, and with only 1.19% of the baseline's computation time. Ablation studies highlight the potential of explicitly designing learned emulators for parameter estimation by leveraging contrastive learning.

TL;DR: This paper describes a novel contrastive framework for learning feature embeddings of observed dynamics jointly with an emulator that can replace high-cost simulators for parameter estimation.

Community Implementations: [ 1 code implementation](https://www.catalyzex.com/paper/arxiv:2211.01554/code)