Optimal Aggregation Strategies for Social Learning Over Graphs

Adaptive social learning is a useful tool for studying distributed decision-making problems over graphs. This paper investigates the effect of combination policies on performance adaptive strategies. Using large-deviation analysis, it first derives bound steady-state error probability and characterizes optimal selection Perron eigenvectors policies. It subsequently studies policy transient behavior strategy by estimating adaptation time in low signal-to-noise ratio regime. In process, discovered that, interestingly, influence insignificant, thus more critical to employ that enhance performance. The theoretical conclusions are illustrated means computer simulations.