Authors: Stephen R. Taylor, Jonathan R. Gair, L. Lentati
Date: 12 Oct 2012
Abstract: We provide brief notes on a particle swarm-optimisation approach to constraining the properties of a stochastic gravitational-wave background in the first International Pulsar Timing Array data-challenge. The technique employs many computational-agents which explore parameter space, remembering their most optimal positions and also sharing this information with all other agents. It is this sharing of information which accelerates the convergence of all agents to the global best-fit location in a very short number of iterations. Error estimates can also be provided by fitting a multivariate Gaussian to the recorded fitness of all visited points.
© M. Vallisneri 2012 — last modified on 2010/01/29
Tantum in modicis, quantum in maximis