We present an algorithm that computes exact maximum flows and minimum-cost flows on directed graphs with m edges and polynomially bounded integral demands, costs, and capacities in m1+o(1) time. Our algorithm builds the flow through a sequence of m1+o(1) approximate undirected minimum-ratio cycles, each of which is computed and processed in amortized mo(1) time using a new dynamic graph data structure.
Our framework extends to algorithms running in m1+o(1) time for computing flows that minimize general edge-separable convex functions to high accuracy. This gives almost-linear time algorithms for several problems including entropy-regularized optimal transport, matrix scaling, p-norm flows, and p-norm isotonic regression on arbitrary directed acyclic graphs.
The maximum flow problem and its generalization, the minimum-cost flow problem, are classic combinatorial graph problems that find numerous applications in engineering and scientific computing. These problems have been studied extensively over the last seven decades, starting from the work of Dantzig and Ford-Fulkerson. Several important algorithmic problems can be reduced to minimum-cost flows, for example, max-weight bipartite matching, min-cut, and Gomory-Hu trees. The origin of numerous significant algorithmic developments such as the simplex method, graph sparsification, and link-cut trees can be traced back to seeking faster algorithms for maximum flow and minimum-cost flow.
1.1. Problem formulation
Formally, we are given a directed graph G = (V, E) with |V| = n vertices and |E| = m edges, upper/lower edge capacities u+, u− ∈ ℝE, edge costs c ∈ ℝE, and vertex demands d ∈ ℝV with ∑v∈V dv = 0. Our goal is to solve the following linear program for the minimum-cost flow problem