Data for "Optimal Static Mutation Strength Distributions for the (1+λ) Evolutionary Algorithm on OneMax"
Maxim Buzdalov, Carola Doerr · Figshare · 2021
This collection contains the data (and the copy of the code that generated that data) for the paper "Optimal Static Mutation Strength Distributions for the (1+λ) Evolutionary Algorithm on OneMax" accepted for the GECCO 2021 conference. The files are: \tone-plus-lambda-on-onemax.tar.gz: the source code to reproduce the experiments. \talgorithm-expectations-flat.csv: the expected running times of various \\((1+\\lambda)\\) algorithms, including the one that uses optimal distributions. \tstatic-N-distributions.csv: the distributions optimized by CMA-ES, where N is the problem size. \tstatic-N-distribution-stddev.csv: maximum deviation between the distributions over the "accepted" runs. \tstatic-N-deviations.csv: the sorted deviations of the outcomes of distribution optimization compared to the best achieved runs. \tstatic-N-summary.csv: the expected running times of the \\((1+\\lambda)\\) EA using the optimal distribution. \tstatic-N-fitness-log.csv: fitness traces from the optimization runs.