Multi-Source Knowledge-Based Hybrid Neural Framework for Time Series Representation Learning

TL;DR

This paper introduces a hybrid neural framework that combines domain-specific knowledge and relational structural understanding to improve multivariate time series forecasting, capturing complex dependencies and uncertainty.

Contribution

It proposes a novel hybrid architecture that integrates domain knowledge with implicit relational structures for enhanced time series representation learning.

Findings

Outperforms state-of-the-art forecasting methods on benchmark datasets.

Effectively models time-varying uncertainty in multi-horizon forecasts.

Demonstrates the importance of combining domain knowledge with structural inference.

Abstract

Accurately predicting the behavior of complex dynamical systems, characterized by high-dimensional multivariate time series(MTS) in interconnected sensor networks, is crucial for informed decision-making in various applications to minimize risk. While graph forecasting networks(GFNs) are ideal for forecasting MTS data that exhibit spatio-temporal dependencies, prior works rely solely on the domain-specific knowledge of time-series variables inter-relationships to model the nonlinear dynamics, neglecting inherent relational structural dependencies among the variables within the MTS data. In contrast, contemporary works infer relational structures from MTS data but neglect domain-specific knowledge. The proposed hybrid architecture addresses these limitations by combining both domain-specific knowledge and implicit knowledge of the relational structure underlying the MTS data using Knowledge-Based Compositional Generalization. The hybrid architecture shows promising results on multiple benchmark datasets, outperforming state-of-the-art forecasting methods. Additionally, the architecture models the time varying uncertainty of multi-horizon forecasts.