Back in early September I started working for Google in Boulder, Colorado. This was quite the switch for me as I’d always assumed I would stay in academia. But they made me an offer I couldn’t refuse, and with Alice turning 5 months old this week its nice to have that extra income and a 9 to 5 job.
My nomination streak continues, this time at the European Conference on Evolutionary Computation in Combinatorial Optimisation (EvoCOP). My paper is titled Hyperplane Elimination for Quickly Enumerating Local Optima. The basic idea is that for certain types of problems, it is possible to determine that large portions of the search space cannot contain any local optima. As a result, Hyperplane elimination was able to find all local optima of landscapes that would take previous methods an estimated 900 million years to complete in just 6 minutes.
Starting January 1st I will be working for Darrell Whitley in the Department of Computer Science at Colorado State University. I’m very excited about this collaboration as much of my recent work has heavily cited Darrell’s.
This year my paper Gray-Box Optimization using the Parameter-less Population Pyramid was nominated for Best Paper in the Genetic Algorithms (GA) track of the Genetic and Evolutionary Computation Conference (GECCO). As usual, all of my code and data is available on github. In order to be nominated, a paper must first be accepted by the track. This year, of the 45 submitted papers, the GA track only accepted 17. During the peer review process, any of the reviewers can suggest a paper for nomination. These suggested papers are then reviewed by the track chairs, and 3 are selected for nomination. This makes my second nomination for best paper, the first being my Parameter-less Population Pyramid paper which won last year’s award.
HeuristicLab is an open source toolkit for performing optimization. Developed by the Heuristic and Evolutionary Algorithms Laboratory (HEAL), this user friendly software provides a clean graphical interface and a heap of features. I am excited to announce that P3 is now available for download as part of HeuristicLab 3.3.11.