Hypervolume approximation

Version 1.0.0.0 (1,32 ko) par Simone
Hypervolume indicator of a Pareto frontier.
641 téléchargements
Mise à jour 15 avr. 2015

Afficher la licence

Approximates the hypervolume of a Pareto frontier. First, it generates
random samples in the hypercuboid defined by the utopia and antiutopia
points. Second, it counts the number of samples dominated by the front.
The hypervolume is approximated as the ratio 'dominated points / total
points'.
Please notice that the choice of the utopia and antiutopia point is
crucial: using points very far from the frontier will result in similar
hypervolume even for very different frontiers (if the utopia is too far
away, the hypervolume will be always low; if the antiutopia is too far
away, the hypervolume will be always high).
Also, frontier points "beyond" the reference points will not be counted
for the approximation (e.g., if the antiutopia is above the frontier or
the utopia is below, the hypervolume will be 0).
Inputs:
- F : the Pareto front to evaluate
- AU : antiutopia point
- U : utopia point
- N : number of sample for the approximation

Outputs:
- hv : hypervolume

Citation pour cette source

Simone (2024). Hypervolume approximation (https://www.mathworks.com/matlabcentral/fileexchange/50517-hypervolume-approximation), MATLAB Central File Exchange. Récupéré le .

Compatibilité avec les versions de MATLAB
Créé avec R2012b
Compatible avec toutes les versions
Plateformes compatibles
Windows macOS Linux
Catégories
En savoir plus sur Multiobjective Optimization dans Help Center et MATLAB Answers
Remerciements

Inspiré par : Hypervolume Indicator

A inspiré : Hypervolume Approximation (MEX)

Community Treasure Hunt

Find the treasures in MATLAB Central and discover how the community can help you!

Start Hunting!
Version Publié le Notes de version
1.0.0.0