Results of the GECCO'2019 1-OBJ Track
"Raw" Result Data
On each problem participants were judged by the best (lowest) function value achieved within the given budget of function evaluations. There were 10 participants in the field. The best function value per problem and participant (1000 times 10 double precision numbers) is listed in this text file.
Participants were ranked based on aggregated problem-wise ranks (details here and here). The following results table lists participants with overall scores (higher is better) and the sum of ranks over all problems (lower is better) The table can be sorted w.r.t. these criteria.
|rank||participant||method name||method description||software||paper||score||sum of ranks|
|1||Nacim Belkhir||AS-AC-CMA-ES||Per Instance Algorithm Selection for low dimension and Per Instance Algorithm Configuration of CMA-ES based on problem features extracted from a uniform sample of the objective function.||link||795.777||3345|
|2||avaneev||biteopt2018||Plain BiteOpt on dimensions < 8, BiteOptDeep on dimensions >= 8.||link||link||651.518||3704|
|6||mini-mlog||GAPSO||A hybrid PSO+DE+Square function model+Polynomial function model with adapted behavior pool and adapted reset behaviors||link||392.868||4523|
|7||Raphael Patrick Prager||166.69||8178|
|8||GERAD||MADS||Mesh Adaptive Direct Search algorithm
Implementation: NOMAD solver
Version: 4.0 (alpha, not available yet)
Visualization of Performance Data
The following figure shows an aggregated view on the performance data.
The following figures show the same data, but separately for each problem dimension.