Funded by the European Union. Views and opinions expressed are however those of the author(s) only and do not necessarily reflect those of the European Union or the European Education and Culture Executive Agency (EACEA). Neither the European Union nor EACEA can be held responsible for them.


On Local Convergence of Stochastic Global Optimization Algorithms

The use of population type nature inspired global optimization algorithms is quite popular due to their attractive natural interpretation and the easy availability of codes. The guarantee of convergence is however doubtful. What this paper investigates is the question how some popular algorithms behave on converging "in the end" to a local minimum point varying dimension and condition of the related Hessian.
The most counterintuitive result is that some popular codes actually do very bad for simple problems.
Basically, this paper calls for investigating algorithms in a systematic way finding out for which cases algorithms do yes or no work well.

Scientific Paper


Target Group:


Cite as:
Hendrix, E.M.T. and Rocha, A.M.A.C. (2021), On Local Convergence of Stochastic Global Optimization Algorithms, O. Gervasi et al. (Eds.): ICCSA Cagliari, Sept. 13-16 2021, LNCS 12953, Springer, Cham, pp. 456-472, doi: 10.1007/978-3-030-86976-2_31

Author of the review:
Eligius Hendrix
University of Malaga


You have to login to leave a comment. If you are not registered click here