The practical robustness of time-series forecasting models on imperfect sensor data remains a critical issue. This paper challenges common heuristics by rigorously comparing two paradigms for achieving robustness: the complex architectural path (Transformer) and the simpler feature engineering path (XGBoost). Using matched hyperparameter search budgets to ensure a fair comparison, our results illuminate a fundamental trade-off between performance consistency and computational efficiency. The well-tuned Modern Transformer achieves the most consistent performance, resulting in the best average MAE of 0.677. In contrast, the feature-based pipelines perform exceptionally well on most data, but are susceptible to a distinct high-error failure mode, resulting in a mean MAE of approximately 0.78. Crucially, we find that this small, statistically insignificant difference in average performance comes at a steep price: the feature-based path is orders of magnitude more computationally efficient. We conclude that the choice of paradigm is not about universal superiority, but a strategic decision. While well-tuned architectures offer the most consistent path to state-of-the-art accuracy, the feature engineering paradigm provides a vastly more efficient baseline. This work provides an empirically-grounded guide to navigating the critical trade-offs between performance and efficiency for real-world forecasting systems.
Ehmimed, N, Chkouri, MY & Touhafi, A 2026, Divergent Pathways to Robustness: On the Trade-offs Between Information Constraint and Architectural Complexity for Imperfect Sensor Data. in 2026 International Conference on Intelligent Systems and Digital Applications (ISDA). IEEE, pp. 1-6. https://doi.org/10.1109/isda70544.2026.11606081
Ehmimed, N., Chkouri, M. Y., & Touhafi, A. (2026). Divergent Pathways to Robustness: On the Trade-offs Between Information Constraint and Architectural Complexity for Imperfect Sensor Data. In 2026 International Conference on Intelligent Systems and Digital Applications (ISDA) (pp. 1-6). IEEE. https://doi.org/10.1109/isda70544.2026.11606081
@inproceedings{c0b6b7d142fc4e83b5e80ef20f0d1918,
title = "Divergent Pathways to Robustness: On the Trade-offs Between Information Constraint and Architectural Complexity for Imperfect Sensor Data",
abstract = "The practical robustness of time-series forecasting models on imperfect sensor data remains a critical issue. This paper challenges common heuristics by rigorously comparing two paradigms for achieving robustness: the complex architectural path (Transformer) and the simpler feature engineering path (XGBoost). Using matched hyperparameter search budgets to ensure a fair comparison, our results illuminate a fundamental trade-off between performance consistency and computational efficiency. The well-tuned Modern Transformer achieves the most consistent performance, resulting in the best average MAE of 0.677. In contrast, the feature-based pipelines perform exceptionally well on most data, but are susceptible to a distinct high-error failure mode, resulting in a mean MAE of approximately 0.78. Crucially, we find that this small, statistically insignificant difference in average performance comes at a steep price: the feature-based path is orders of magnitude more computationally efficient. We conclude that the choice of paradigm is not about universal superiority, but a strategic decision. While well-tuned architectures offer the most consistent path to state-of-the-art accuracy, the feature engineering paradigm provides a vastly more efficient baseline. This work provides an empirically-grounded guide to navigating the critical trade-offs between performance and efficiency for real-world forecasting systems.",
author = "Nadir Ehmimed and Chkouri, \{Mohamed Yassin\} and Abdellah Touhafi",
note = "Publisher Copyright: {\textcopyright} 2026 IEEE.",
year = "2026",
month = may,
doi = "10.1109/isda70544.2026.11606081",
language = "English",
pages = "1--6",
booktitle = "2026 International Conference on Intelligent Systems and Digital Applications (ISDA)",
publisher = "IEEE",
}