A discrete Benamou-Brenier formulation of Optimal Transport on graphs
By: Kieran Morris, Oliver Johnson
Potential Business Impact:
Finds shortest paths on networks for data.
We propose a discrete transport equation on graphs which connects distributions on both vertices and edges. We then derive a discrete analogue of the Benamou-Brenier formulation for Wasserstein-$1$ distance on a graph and as a result classify all $W_1$ geodesics on graphs.
Similar Papers
Dynamic characterization of barycentric optimal transport problems and their martingale relaxation
Probability
Finds new ways to move data smoothly.
Vector valued optimal transport: from dynamic to static formulations
Analysis of PDEs
Unifies math for complex problems, speeding up calculations.
On a $T_1$ Transport inequality for the adapted Wasserstein distance
Probability
Helps computers understand how things change over time.