Skip to content

RachM/length-scale

Repository files navigation

Analysing and Characterising Optimization Problems Using Length Scale

Source for calculating the length scales and Dj between BBOB function instances. Used in the experiments in [1].

To use:

calculateCorrelationWithERT.m:

  • Script for calculating the correlation between each metric and the best ERT

calculateDistributionsAndKL.m:

  • Script for calculating the length scale distributions and KL divergence

calculateDistributionsAndKLBetweenD.m:

  • Calculates the length scale distributions and KL divergence between BBOB problems of different dimensions

calculateMetrics.m:

  • Script for calculating landscape metrics of the BBOB problems

plotDistributionsBetweenD.m:

  • Script for plotting the length scale distributions across dimensionality

plotDjResults.m:

  • Script for plotting the J-divergence visualisations, heatmaps and dendrograms

plotFeatureVectorSpace.m:

  • Script for plotting the visualisations, heatmaps and dendrograms of the landscape feature feature vector

plotMetricsVsERT.m:

  • Script for plotting landscape metrics against the best ERT

sampleBBOB.m:

  • Script for sampling the BBOB functions

[1] R. Morgan and M. Gallagher. Analysing and Characterising Optimization Problems Using Length Scale. Soft Computing, 2015 (accepted).

About

Matlab code for length scale analysis

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published

Languages