Meal-induced glucose instability is detected from 15-minute windows of CGM data using DMD eigenvalues, with a simulator-trained logistic regression that outperforms a clinical baseline on two real datasets.
Title resolution pending
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.LG 1years
2025 1verdicts
CONDITIONAL 1representative citing papers
citing papers explorer
-
Online Meal Detection Based on CGM Data Dynamics
Meal-induced glucose instability is detected from 15-minute windows of CGM data using DMD eigenvalues, with a simulator-trained logistic regression that outperforms a clinical baseline on two real datasets.