English

Hybrid Meta-Learning Framework for Anomaly Forecasting in Nonlinear Dynamical Systems via Physics-Inspired Simulation and Deep Ensembles

Machine Learning 2025-06-18 v1 Systems and Control Systems and Control Accelerator Physics

Abstract

We propose a hybrid meta-learning framework for forecasting and anomaly detection in nonlinear dynamical systems characterized by nonstationary and stochastic behavior. The approach integrates a physics-inspired simulator that captures nonlinear growth-relaxation dynamics with random perturbations, representative of many complex physical, industrial, and cyber-physical systems. We use CNN-LSTM architectures for spatio-temporal feature extraction, Variational Autoencoders (VAE) for unsupervised anomaly scoring, and Isolation Forests for residual-based outlier detection in addition to a Dual-Stage Attention Recurrent Neural Network (DA-RNN) for one-step forecasting on top of the generated simulation data. To create composite anomaly forecasts, these models are combined using a meta-learner that combines forecasting outputs, reconstruction errors, and residual scores. The hybrid ensemble performs better than standalone models in anomaly localization, generalization, and robustness to nonlinear deviations, according to simulation-based experiments. The framework provides a broad, data-driven approach to early defect identification and predictive monitoring in nonlinear systems, which may be applied to a variety of scenarios where complete physical models might not be accessible.

Keywords

Cite

@article{arxiv.2506.13828,
  title  = {Hybrid Meta-Learning Framework for Anomaly Forecasting in Nonlinear Dynamical Systems via Physics-Inspired Simulation and Deep Ensembles},
  author = {Abdullah Burkan Bereketoglu},
  journal= {arXiv preprint arXiv:2506.13828},
  year   = {2025}
}

Comments

6 pages, 5 figures, 5 algorithms

R2 v1 2026-07-01T03:20:22.870Z