Sunrise atop Golden Mountain: the Angkor ruins of Cambodia

Home

             Welcome to my website.  For thesis, I visualized the occurrence of premature convergence in particle swarm optimization (PSO), understood the value of monitoring the source of the problem in order to detect its occurrence efficiently rather than monitoring function values to detect it indirectly, experimentally confirmed the hypothesis that direct detection of premature convergence works better than indirect detection, conceptualized an efficient regrouping mechanism based on the swarm state at premature convergence, wrote and tested the regrouping algorithm, and fine-tuned the process resulting in Regrouping PSO (RegPSO) published in thesis and presented at the 2009 IEEE International Conference on Systems, Man, and Cybernetics.
             At the same time, thousands of parameter combinations were empirically tested in order to see how well the stagnation problem could be combated without adding a novel regrouping mechanism.  Results were surprisingly good but hinted at greater problem dependence than produced by regrouping the swarm.
             Legal disclaimer: the content of this website is protected by copyright laws and should not be misunderstood to be in the public domain.  To cite any information on this website, please cite thesis or the website itself.

To contact me:
E-mail: george at georgeevers.org

(@ removed to avoid email harvesters)

George Evers, MSE

Electrical Engineer with Programming Emphasis