Online Learning via Learned Latent Bayesian Tracking

2026-09-25

AI Summary

This study addresses the computational bottlenecks of Bayesian filtering and the lack of low-dimensional dynamic representations in online learning for high-dimensional models by proposing the AURA framework. To the best of our knowledge, this work is the first to introduce an adaptation-aware latent geometry, thereby eliminating reliance on handcrafted subspaces. Specifically, AURA learns a latent state-space model offline and subsequently employs an extended Kalman filter online to efficiently update parameters within the low-dimensional space before reconstructing the full model. Evaluated on wireless channel estimation and non-stationary image classification tasks, AURA significantly enhances adaptation speed, predictive accuracy, and computational efficiency, consistently outperforming existing baseline methods across all metrics.

Abstract

Online learning in non-stationary environments requires models to adapt rapidly from streaming data under strict computational constraints. A principled approach casts online learning as Bayesian state tracking, where model parameters are updated sequentially via Bayesian filtering. However, applying Bayesian filters directly to modern deep models is computationally prohibitive due to the high dimensionality of parameter space, forcing existing methods to rely on restrictive approximations or manually designed low-dimensional subspaces. In this work, we identify the absence of a suitable low-dimensional dynamical representation as the core bottleneck in Bayesian filtering-based online learning. Accordingly, we propose Adaptive Update through Representation Adaptation (AURA), a meta-learning framework that learns offline a low-dimensional latent state-space model governing the evolution of optimal model parameters under distribution shift. Online adaptation is then performed via extended Kalman filtering in this learned latent space followed by reconstruction of the full model parameters through a learned lifting map, enabling efficient single-step online adaptation while preserving model expressiveness. Evaluated on online adaptation of neural wireless receivers under time-varying channels and on non-stationary image classification, AURA shows substantial improvements in adaptation speed, accuracy, and computational efficiency over existing online learning and Bayesian filtering baselines, demonstrating that an adaptation-aware latent geometry is beneficial for effective Bayesian online learning in high-dimensional models.

Problem

• online learning
• Bayesian filtering
• non-stationary environments
• high-dimensional parameter space
• latent representation

Innovation

• Online Learning
• Bayesian Filtering
• Meta-Learning
• Latent State-Space Model
• Extended Kalman Filter