Domain Adaptation for Industrial Time-series Forecasting via Counterfactual Inference

TL;DR

This paper introduces a causal domain adaptation framework for industrial time-series forecasting that leverages shared causality and counterfactual inference to improve predictions in data-scarce target domains, aiding industrial decision-making.

Contribution

The paper proposes a novel causal domain adaptation method with an answer-based attention mechanism for counterfactual prediction in industrial time-series, addressing data scarcity and treatment variability.

Findings

Outperforms baseline methods on real-world oilfield datasets

Effectively models treatments and outcomes jointly across domains

Provides accurate counterfactual predictions to guide production processes

Abstract

Industrial time-series, as a structural data responds to production process information, can be utilized to perform data-driven decision-making for effective monitoring of industrial production process. However, there are some challenges for time-series forecasting in industry, e.g., predicting few-shot caused by data shortage, and decision-confusing caused by unknown treatment policy. To cope with the problems, we propose a novel causal domain adaptation framework, Causal Domain Adaptation (CDA) forecaster to improve the performance on the interested domain with limited data (target). Firstly, we analyze the causality existing along with treatments, and thus ensure the shared causality over time. Subsequently, we propose an answer-based attention mechanism to achieve domain-invariant representation by the shared causality in both domains. Then, a novel domain-adaptation is built to model treatments and outcomes jointly training on source and target domain. The main insights are that our designed answer-based attention mechanism allows the target domain to leverage the existed causality in source time-series even with different treatments, and our forecaster can predict the counterfactual outcome of industrial time-series, meaning a guidance in production process. Compared with commonly baselines, our method on real-world and synthetic oilfield datasets demonstrates the effectiveness in across-domain prediction and the practicality in guiding production process