Closing the loop in learning with missing data
2026-08-10 • Machine Learning
Machine Learning
AI summaryⓘ
The authors study how machine learning models should learn when some data is missing during training. They use ideas from dynamical systems to view missing data as a limitation that affects how accurately the model parameters can be updated. Their method adapts model updates to keep the learning stable and consistent even when some information is missing or only occasionally available. They test their approach on data with multiple types of inputs and find it helps maintain learning stability in very sparse data situations.
machine learningmissing datadynamical systemscontrollabilityLyapunov stabilityadaptive learningobservabilitymultimodal datastability analysis
Authors
Dimitrios Pylorof, Humberto E. Garcia
Abstract
What should a machine learning model learn when data is missing during training? We look at the learning process from a dynamical systems perspective, cast data missingness as a structured loss of actuation that limits controllability of the parameter error dynamics, and ultimately derive adaptation mechanisms with Lyapunov stability characteristics that throttle model updates in ways that preserve learning coherence under partial, intermittent observability. Under recurrent excitation, our analysis provides ISS-type residual-to-state bounds with respect to a bounded closed-loop mismatch between the loss residual and the preconditioned update geometry. We evaluate the efficacy of our directional observability-aware adaptive learning approach on multimodal contexts, reinforcing its premise in promoting learning coherence and stability even in pathologically sparse domains and problems.