Pith. sign in

REVIEW 1 cited by

Learning Mixtures of Linear Dynamical Systems

Not yet reviewed by Pith; the record is open.

This paper has not been read by Pith yet. Machine review is queued; the pith claim, tier, and objections will appear here once it completes.

SPECIMEN: schema-true, not a live event

T0 review · schema-true

One-sentence machine reading of the paper's core claim.

pith:XXXXXXXX · record.json · timestamp

arxiv 2201.11211 v2 pith:MAYB47KJ submitted 2022-01-26 stat.ML cs.LGcs.SYeess.SYmath.STstat.TH

Learning Mixtures of Linear Dynamical Systems

classification stat.ML cs.LGcs.SYeess.SYmath.STstat.TH
keywords learningmodelssampletrajectorieschallengesdatadynamicallinear
verification ladder T0 review T1 audit T2 compute T3 formal T4 reserved
0 comments
read the original abstract

We study the problem of learning a mixture of multiple linear dynamical systems (LDSs) from unlabeled short sample trajectories, each generated by one of the LDS models. Despite the wide applicability of mixture models for time-series data, learning algorithms that come with end-to-end performance guarantees are largely absent from existing literature. There are multiple sources of technical challenges, including but not limited to (1) the presence of latent variables (i.e. the unknown labels of trajectories); (2) the possibility that the sample trajectories might have lengths much smaller than the dimension $d$ of the LDS models; and (3) the complicated temporal dependence inherent to time-series data. To tackle these challenges, we develop a two-stage meta-algorithm, which is guaranteed to efficiently recover each ground-truth LDS model up to error $\tilde{O}(\sqrt{d/T})$, where $T$ is the total sample size. We validate our theoretical studies with numerical experiments, confirming the efficacy of the proposed algorithm.

discussion (0)

Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.

Forward citations

Cited by 1 Pith paper

Reviewed papers in the Pith corpus that reference this work. Sorted by Pith novelty score.

  1. Structure learning of Hamiltonians from real-time evolution

    quant-ph 2024-04 unverdicted novelty 7.0

    New algorithm learns unknown local Hamiltonians from real-time evolution with total time O(log n / ε), without knowing terms, for bounded-norm interactions.