Opened 13 months ago

Last modified 3 months ago

#2592 reviewing feature request

MO-CMA-ES for Realnumbered Problem

Reported by: bwerth Owned by: pfleck
Priority: medium Milestone: HeuristicLab 3.3.15
Component: Algorithms.CMAEvolutionStrategy Version: 3.3.13
Keywords: Cc:

Description

A multiobjective CMA-ES variant for realnumbered multiobjective problems should be created.

Change History (14)

comment:1 Changed 13 months ago by bwerth

r13730 create new branch folder r13731 initial branch r13732 rename folder

comment:2 Changed 12 months ago by bwerth

r13793 first unfinished implementation similiar to Shark – Machine Learning 3.1

comment:3 Changed 10 months ago by bwerth

r13909 added analysiation and CrowdingIndicator

comment:4 Changed 10 months ago by mkommend

  • Milestone changed from HeuristicLab 3.3.14 to HeuristicLab 3.3.15

comment:5 Changed 10 months ago by bwerth

Last edited 10 months ago by bwerth (previous) (diff)

comment:6 Changed 10 months ago by bwerth

r13990 reimplemented Indicators

comment:7 Changed 8 months ago by bwerth

r14269 code cleanup + project refactored

Last edited 8 months ago by bwerth (previous) (diff)

comment:8 Changed 5 months ago by bwerth

  • Status changed from new to assigned

comment:9 Changed 5 months ago by bwerth

  • Owner changed from bwerth to mkommend
  • Status changed from assigned to reviewing

comment:10 Changed 5 months ago by bwerth

r14404 several fixes and cleanups to adapt a more HeuristicLab-Code-Style + renaming of folders and Plugin

comment:11 Changed 5 months ago by mkommend

  • Owner changed from mkommend to pfleck

comment:12 Changed 3 months ago by bwerth

r14577 made MOCMAES compatible with MultiObjectiveBasicProblem instead of MultiObjectiveTestfunction, fixed Bug in CrowdingIndicator

comment:13 Changed 3 months ago by bwerth

r14612 made initialization solutionspace-scaling invariant

comment:14 Changed 3 months ago by pfleck

r14614

  • Fixed scale invariant initialization by changing the initial sigma to a DoubleArray for initializing the covariance matrix (similar to the CMA-ES).
  • Made the MO-CMA-ES pausable (although the algorithm state is not yet storable when paused).
Note: See TracTickets for help on using tickets.