Usporedba neuronskih mreža učenih algoritmima rojeva

master's thesis

• Title (english): Comparison of neural networks learned with swarm algorithms
• Author: Krešimir Milas
• Mentor: Dubravko Majetić (mentor)
• Granter: University of Zagreb; Faculty of Mechanical Engineering and Naval Architecture; Zagreb
• Defense date and country: 2017-01-26, Croatia
• Scientific / art field, discipline and subdiscipline: TECHNICAL SCIENCES; Mechanical Engineering

Abstract

U ovom radu se obrađuju neuronske mreže učene algoritmima rojeva. Koriste se dva najčešće korištena algoritma rojeva: PSO (Particle Swarm Optimization) i ABC(Artificial Bee Colony). Prvo se istražuje utjecaj parametara učenja pojedinih algoritama na kvalitetu i brzinu učenja, a potom se mreže učene različitim algoritmima rojeva uspoređuju međusobno po kriterijima kvalitete i brzine učenja. Uspoređuju se mreže s koje su imale najbolja generalizacijska svojstva. Na kraju se uspoređuju mreže učene algoritmima rojeva s mrežama učenim algoritmom s povratnim prostiranjem greške.

Abstract (english)

This paper dealt with neural networks trained using swarm intelligence. Two most common swarm intelligence algorithms have been used: PSO (Particle Swarm Optimization) and ABC (Artificial Bee Colony). Firstly, the impact of training parameters on training process has been explored and then neural networks trained by different swarm intelligence algorithms have been compared on error achieved and epochs elapsed criteria. Neural networks compared were ones that achieved best generalization performance. Lastly, neural networks trained with swarm intelligence are compared to error-backpropagation neural networks.

• Keywords: neuronske mreže; umjetna inteligencija; algoritmi rojeva; PSO; ABC; linearni dinamički sustav; P1 član; nelinearni dinamički sustav
• Keywords (english): neural networks; swarm intelligence; PSO; ABC; swarm intelligence; linear dynamical system; first order lag element; non-linear dynamical system
• Language: croatian
• URN:NBN: urn:nbn:hr:235:006515
• Study programme: Title: Mechanical Engineering; Study programme type: university; Study level: graduate; Academic / professional title: magistar/magistra inženjer/inženjerka strojarstva (magistar/magistra inženjer/inženjerka strojarstva)
• Type of resource: Text
• File origin: Born digital
• Access conditions: Open access
• Created on: 2019-05-22 18:30:00