A novel fitting algorithm based on Bacterial Swarm Optimizer for stochastic data

This paper proposes a novel stochastic algorithm, which aims to describe the random distributions of experimentally acquired data. Generally, such data can be satisfactorily modeled through the use of a Gaussian distribution. However, it is not always the case, instances can arise in which the distributions of measured data are not strictly Gaussian in their nature. The present work adopts Bacterial Swarm Optimizer (BSO), which has been inspired from bacterial foraging behavior and quorum sensing, to optimize the Probability Density Function (PDF) for describing a particle identification spectrum constructed from data collected in an experiment undertaken at Gesellschaft fur Schwerionenforschung (GSI), Darmstadt, Germany. Our studies indicates that the PDF proposed in the present paper is more accurate than that of several convention methods.