AALF: Almost Always Linear Forecasting

2024-09-16

arXiv.org

AI Summary

Balancing prediction accuracy and interpretability remains challenging in time-series forecasting. Method: This paper proposes a lightweight online model selection framework that defaults to a linear autoregressive (AR) model and dynamically switches to a deep learning model only for a small subset of critical forecast points. It introduces an interpretable, meta-feature–driven decision mechanism with adaptive online thresholds—avoiding opaque, indiscriminate use of black-box models. Contribution/Results: The work provides the first systematic empirical validation that “using linear models almost always” yields competitive forecasting performance. Evaluated across multiple real-world datasets, the framework achieves accuracy on par with state-of-the-art online model selection methods, while significantly improving interpretability, computational efficiency, and deployment robustness. This establishes a new paradigm for high-assurance time-series forecasting.

Technology Category

• Machine Learning: Time-Series/Data Streams
• Natural Language Processing: Interpretability, Analysis, and Evaluation of NLP Models
• Planning, Routing, and Scheduling: Model-Based Reasoning

Application Category

• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking
• User Modeling, Personalization and Recommendation: Explainable and interpretable methods for personalization
• Systems and Infrastructure for Web, Mobile and WoT: Applied ML and AI for Web-based mobile applications

Abstract

Recent works for time-series forecasting more and more leverage the high predictive power of Deep Learning models. With this increase in model complexity, however, comes a lack in understanding of the underlying model decision process, which is problematic for high-stakes application scenarios. At the same time, simple, interpretable forecasting methods such as ARIMA still perform very well, sometimes on-par, with Deep Learning approaches. We argue that simple models are good enough most of the time, and that forecasting performance could be improved by choosing a Deep Learning method only for few, important predictions, increasing the overall interpretability of the forecasting process. In this context, we propose a novel online model selection framework which learns to identify these predictions. An extensive empirical study on various real-world datasets shows that our selection methodology performs comparable to state-of-the-art online model selections methods in most cases while being significantly more interpretable. We find that almost always choosing a simple autoregressive linear model for forecasting results in competitive performance, suggesting that the need for opaque black-box models in time-series forecasting might be smaller than recent works would suggest.

Problem

• Time Series Prediction
• Interpretability
• Model Simplicity

Innovation

• Time Series Prediction
• Selective Deep Learning Application
• Interpretable Forecasting