Randomized Shortest Paths with Net Flows and Capacity Constraints
This work incrementally improves network analysis methods for applications like clustering and routing by addressing flow neutrality and capacity limits.
The authors extended the randomized shortest paths (RSP) model by introducing net flows, where opposite-direction flows cancel out, and capacity constraints on edge flows, developing algorithms for computing node dissimilarities and solving constrained problems. Experimental results showed that the net flow RSP dissimilarity is competitive with state-of-the-art methods in node clustering tasks.
This work extends the randomized shortest paths (RSP) model by investigating the net flow RSP and adding capacity constraints on edge flows. The standard RSP is a model of movement, or spread, through a network interpolating between a random-walk and a shortest-path behavior [30, 42, 49]. The framework assumes a unit flow injected into a source node and collected from a target node with flows minimizing the expected transportation cost, together with a relative entropy regularization term. In this context, the present work first develops the net flow RSP model considering that edge flows in opposite directions neutralize each other (as in electric networks), and proposes an algorithm for computing the expected routing costs between all pairs of nodes. This quantity is called the net flow RSP dissimilarity measure between nodes. Experimental comparisons on node clustering tasks indicate that the net flow RSP dissimilarity is competitive with other state-of-the-art dissimilarities. In the second part of the paper, it is shown how to introduce capacity constraints on edge flows, and a procedure is developed to solve this constrained problem by exploiting Lagrangian duality. These two extensions should improve significantly the scope of applications of the RSP framework.