Fig. 2: Performance benchmark.
From: A collective intelligence model for swarm robotics applications
Histogram matrices of success rate SR (a) and mean number of function evaluations 〈FE〉 (b). Each row corresponds to a landscape function (the reader is referred to section 4 of SI for landscape formulas), whereas each column corresponds to a dimension, N. The single histogram compares SR or 〈FE〉 of three methods, Particle Swarm Optimization (PSO), Multistart Interior-Point Algorithm (MIPA) and Swarm Cooperation Model (SCM), for different swarm sizes, M. The box plots about 〈FE〉 indicates the mean value and the corresponding standard deviation.
