Performance Measurement for Interactive Multi-objective Evolutionary Algorithms

This paper suggests to use a different metric for performance of multiple-point interactive evolutionary multi-objective algorithms. We defined a preferred region based on a set of user's reference points. Based on the preferred region, we also define a User based Front (UbF) which is generated...

Celý popis

Uloženo v:
Podrobná bibliografie
Vydáno v:2015 Seventh International Conference on Knowledge and Systems Engineering (KSE) s. 302 - 305
Hlavní autoři: Long Nguyen, Hung Nguyen Xuan, Lam Thu Bui
Médium: Konferenční příspěvek
Jazyk:angličtina
Vydáno: IEEE 01.10.2015
Témata:
Convergence
DMEA-II
Evolutionary computation
IGD
Interactive EMO
Measurement
MOEA/D
multi-point interactive
Optical fibers
Optimization
Reference points
Sociology
Statistics
On-line přístup:Získat plný text
Tagy: Přidat tag
Žádné tagy, Buďte první, kdo vytvoří štítek k tomuto záznamu!
Popis
Shrnutí:This paper suggests to use a different metric for performance of multiple-point interactive evolutionary multi-objective algorithms. We defined a preferred region based on a set of user's reference points. Based on the preferred region, we also define a User based Front (UbF) which is generated from the preferred region. UbF is used in calculation of Generational Distance (GD) and Inverse Generational Distance (IGD). The usage of the metric in experiments indicated meaningful comparisons for interactive multi-objective evolutionary algorithms using multiple reference points.
DOI:10.1109/KSE.2015.51