A unified online algorithm predicts any LDS with Õ(k) parameters (k = instability complexity), matching lower bound, and beats equal-budget baselines on high-d systems.
Title resolution pending
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.LG 1years
2026 1verdicts
ACCEPT 1representative citing papers
citing papers explorer
-
A Memory Efficient Unified Algorithm for Online Learning of Linear Dynamical Systems
A unified online algorithm predicts any LDS with Õ(k) parameters (k = instability complexity), matching lower bound, and beats equal-budget baselines on high-d systems.