Cost-based analyses of random neighbor and derived sampling methods

Abstract Random neighbor sampling, or RN , is a method for sampling vertices with mean degree greater than that of the graph. Instead naïvely vertex from graph and retaining it (‘random vertex’ RV ), selected instead. While considerable research has analyzed various aspects extra cost second typically not addressed. This paper explores perspective cost. We break down into two distinct costs, an already sampled vertex, we also include actually selecting vertex/neighbor use rather discarding it. With these three costs as our cost-model, explore compare to in more fair manner comparisons have been made previous research. As delve number variants are introduced. These improve on cost-effectiveness regard particular priorities. Our full cost-benefit analysis highlights strengths weaknesses methods. particularly focus how methods perform high-degree low-degree vertices, which further enriches understanding they can be practically applied. suggest ‘two-phase’ specifically seek cover both separate phases.