Prilagodljivi algoritem diferencialne evolucije z arhivom uspešnosti in linearnim zmanjševanjem populacije

diplomsko delo

Aleš Gartner (Author), Iztok Fister (Mentor), Iztok Fister (Co-mentor)

Abstract

V sklopu diplomskega dela predstavljamo delovanje prilagodljivega algoritma diferencialne evolucije z arhivom uspešnosti in linearnim zmanjševanjem populacije ter ga implementiramo v programskem jeziku Python. S statistično primerjavo rezultatov implementacije na testnih funkcijah smo pokazali, da smo algoritem uspešno implementirali. Algoritem smo vključili v Python knjižnico NiaPy ter primerjali njegovo učinkovitost z drugimi algoritmi diferencialne evolucije, implementiranimi v NiaPy. Z analizo rezultatov smo pokazali, da je naš implementirani algoritem resnično eden izmed najučinkovitejših verzij algoritma diferencialne evolucije.

Keywords

optimizacija;algoritmi po vzoru iz narave;diferencialna evolucija;diplomske naloge;

Data

Language:	Slovenian
Year of publishing:	2022
Typology:	2.11 - Undergraduate Thesis
Organization:	UM FERI - Faculty of Electrical Engineering and Computer Science
Publisher:	[A. Gartner]
UDC:	004.8.021(043.2)
COBISS:	137305859
Views:	93
Downloads:	8
Average score:	0 (0 votes)
Metadata:

Other data

Secondary language:	English
Secondary title:	Succes-History based adaptive differential evolution algorithm with linear population size reduction
Secondary abstract:	As part of our thesis, we have presented the operation of the Success-History based Adaptive Differential Evolution algorithm with Linear Population Size Reduction and implemented it in the Python programming language. Through statistical comparison of the results on test functions, we have demonstrated that the algorithm was successfully implemented. We merged the algorithm into the Python library NiaPy and compared its performance with other already existing differential evolution algorithms implemented in the same library. By analysing the results, we justify that our implemented algorithm is among the best preforming variants of the differential evolution algorithm.
Secondary keywords:	optimization;nature-inspired algorithms;differential evolution;NiaPy;
Type (COBISS):	Bachelor thesis/paper
Thesis comment:	Univ. v Mariboru, Fak. za elektrotehniko, računalništvo in informatiko, Računalništvo in informacijske tehnologije
Pages:	1 spletni vir (1 datoteka PDF (X, 29 f.))
ID:	16431785