Mining Causality from Continuous-time Dynamics Models: An Application to Tsunami Forecasting

TL;DR

This paper introduces a method to extract causal relationships from neural continuous-time models, specifically applied to tsunami forecasting, improving interpretability without sacrificing forecasting accuracy.

Contribution

It proposes a sparsity-based mechanism to mine causal structures from neural ODE models, bridging the gap between accurate predictions and causal interpretability.

Findings

Effective in recovering known causal structures in synthetic data.

Successfully applied to tsunami forecasting with high accuracy.

Learns physically consistent causal relationships.

Abstract

Continuous-time dynamics models, such as neural ordinary differential equations, have enabled the modeling of underlying dynamics in time-series data and accurate forecasting. However, parameterization of dynamics using a neural network makes it difficult for humans to identify causal structures in the data. In consequence, this opaqueness hinders the use of these models in the domains where capturing causal relationships carries the same importance as accurate predictions, e.g., tsunami forecasting. In this paper, we address this challenge by proposing a mechanism for mining causal structures from continuous-time models. We train models to capture the causal structure by enforcing sparsity in the weights of the input layers of the dynamics models. We first verify the effectiveness of our method in the scenario where the exact causal-structures of time-series are known as a priori. We next apply our method to a real-world problem, namely tsunami forecasting, where the exact causal-structures are difficult to characterize. Experimental results show that the proposed method is effective in learning physically-consistent causal relationships while achieving high forecasting accuracy.