Math 2.0

Resources / preprint

Prediction Limits and Koopman Closure of Geometry-Induced Soft State Abstractions 6KAZWJ8RFX

URIm2://res/6KAZWJ8RFX
Kindpreprint
CreatorsMohit Kumar, Somayeh Kargaran
Date2026-09-26
Subjectsarxiv cs.AI, arxiv cs.LG, arxiv math.DS
Scopeborder

Expressions

e1: version, preprint

Date2026-09-26
IdsarXiv 2609.32652

Items in this resource

The mentions of the inventory that name this resource. Each name links to the card of its inventory row.

ItemKindWhereSource
Koopman candidate residual boundstatementcorollary 5.15arxiv
Adjoint Koopman Eigenpairdefinitiondefinition 3.4arxiv
Finite-sample class-centered prediction-error certificatestatementproposition 2.3arxiv
Squared Relative Adjoint Residualdefinitiondefinition 4.10arxiv
Squared deviation from reference labelsdefinitiondefinition 5.3arxiv
Sensitivity to reference-label deviationstatementproposition 5.5arxiv
Class occupancy and downward bias of within-class sample variancesstatementproposition 2.5arxiv
Certificate using prespecified class representativesstatementcorollary 2.4arxiv
80JRVQ1ZVDstatementproposition 4.7arxiv
Exact NLMS parameter-error identitystatementproposition 5.1arxiv
Risk bound from independent evaluation pairsstatementproposition 5.13arxiv
Spectral Correspondence under Exact Closurestatementproposition 4.9arxiv
Tolerance for the squared relative adjoint residualdefinitiondefinition 4.11arxiv
Compact one-step closure risk boundstatementtheorem 5.12arxiv
Attainable error with bounded matrix normstatementproposition 5.9arxiv
Reconstruction-based sufficient conditions for a marginstatementcorollary 5.7arxiv
Spectral Evaluation Datasetdefinitiondefinition 4.3arxiv
Bound on deviation from one-hot reference labelsstatementcorollary 5.16arxiv
Complexified RKHSdefinitiondefinition 3.3arxiv
Training transitions and reference datasetsdefinitiondefinition 4.1arxiv
Consistency and detection above a fixed error tolerancestatementcorollary 2.6arxiv
Projection onto the span of feature vectorsstatementproposition 4.8arxiv
Family of KAHM Soft Abstractionsdefinitiondefinition 4.4arxiv
Closure-risk bounds with a finite-sample assignment termstatementtheorem 5.11arxiv
Reconstruction-score margins bound deviation from reference labelsstatementtheorem 5.6arxiv
Fitted linear coordinate modeldefinitiondefinition 4.15arxiv
NVTT3B98SDstatementproposition 3.2arxiv
$t$-Step Evaluation Datasetdefinitiondefinition 4.2arxiv
Class-centered population prediction-risk boundstatementproposition 2.1arxiv
Assignment error, coordinate second moments, and approximation errorstatementtheorem 5.4arxiv
Upper bounds on one-step prediction errorstatementproposition 5.2arxiv
KAHM reconstruction scoredefinitiondefinition 3.1arxiv
Linear prediction for a fixed soft abstractionproblemproblem 4.13arxiv
Computable prediction-error bound under a spectral-norm limitstatementproposition 2.2arxiv

Catalogue records