DICE recovers the vector field driving population-level dynamics from unpaired time marginals through a discrete inverse continuity equation, with a unique minimizer and a first-order time-discretization error.
Gradient Flows
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
-
DICE: Discrete inverse continuity equation for learning population dynamics
DICE recovers the vector field driving population-level dynamics from unpaired time marginals through a discrete inverse continuity equation, with a unique minimizer and a first-order time-discretization error.