![[2602.22294] When Should a Model Change Its Mind? An Energy-Based Theory and Regularizer for Concept Drift in Electrocardiogram (ECG) Signals](https://arxiv.org/static/browse/0.3.4/images/arxiv-logo-fb.png){kind=logo}
[2602.22294] When Should a Model Change Its Mind? An Energy-Based Theory and Regularizer for Concept Drift in Electrocardiogram (ECG) Signals
Summary
This paper presents an energy-based framework for managing concept drift in ECG signals, proposing a new regularizer that enhances model stability without altering architectures.
Why It Matters
Understanding and addressing concept drift is crucial in medical signal processing, particularly for ECG analysis. This research introduces a novel approach that could improve the reliability of machine learning models in dynamic environments, ultimately enhancing patient monitoring and diagnosis.
Key Takeaways
- Introduces Physiologic Energy Conservation Theory (PECT) for concept stability.
- Proposes Energy-Constrained Representation Learning (ECRL) to manage latent movement.
- Demonstrates improved accuracy in ECG models under perturbed conditions.
- Provides empirical evidence supporting the energy-drift law for dynamic signals.
- Enhances understanding of benign variability versus true concept change.
Computer Science > Machine Learning arXiv:2602.22294 (cs) [Submitted on 25 Feb 2026] Title:When Should a Model Change Its Mind? An Energy-Based Theory and Regularizer for Concept Drift in Electrocardiogram (ECG) Signals Authors:Timothy Oladunni, Blessing Ojeme, Kyndal Maclin, Clyde Baidoo View a PDF of the paper titled When Should a Model Change Its Mind? An Energy-Based Theory and Regularizer for Concept Drift in Electrocardiogram (ECG) Signals, by Timothy Oladunni and 3 other authors View PDF HTML (experimental) Abstract:Models operating on dynamic physiologic signals must distinguish benign, label-preserving variability from true concept change. Existing concept-drift frameworks are largely distributional and provide no principled guidance on how much a model's internal representation may move when the underlying signal undergoes physiologically plausible fluctuations in energy. As a result, deep models often misinterpret harmless changes in amplitude, rate, or morphology as concept drift, yielding unstable predictions, particularly in multimodal fusion settings. This study introduces Physiologic Energy Conservation Theory (PECT), an energy-based framework for concept stability in dynamic signals. PECT posits that under virtual drift, normalized latent displacement should scale proportionally with normalized signal energy change, while persistent violations of this proportionality indicate real concept drift. We operationalize this principle through Energy-Constrained R...
