Primjena metoda strojnog učenja u proizvodnim sustavima

De Marco, Marjam (2018) Primjena metoda strojnog učenja u proizvodnim sustavima. = Application of machine learning methods in production systems. Master's thesis (Bologna) , Sveučilište u Zagrebu, Fakultet strojarstva i brodogradnje, UNSPECIFIED. Mentor: Lisjak, Dragutin.

Preview	Text De_Marco_2018_diplomski.pdf - Published Version Jezik dokumenta:Croatian Download (4MB) | Preview
	Text De Marco_Marjam_autorska_izjava_diplomski_2018.pdf - Published Version Restricted to Repository staff only Jezik dokumenta:Croatian Download (517kB)

Abstract (Croatian)

Tema ovog diplomskog rada je primjena metoda strojnog učenja u linijskoj proizvodnji protupožarnih zaklopki i regulatora varijabilnog protoka za operaciju montaže. Primjena strojnog učenja i dubinske analize podataka postaje standard u svim aspektima proizvodnje, kako bi se otkrile skrivene informacije i znanje utkano u podacima, a uvelike pridonoseći procesu donošenja odluka i poslovanja. Cilj ovog rada je pronaći matematički model koji će prema kriteriju točnosti analizirati podatke te doprinijeti razumijevanju procesa montaže linijske proizvodnje. U prvom djelu rada opisana je teorija strojnog učenja, tehnike za dubinsku analizu podataka – klaster analiza, regresija i klasifikacija. U praktičnom dijelu rada, na podacima iz poslovnog sustava napravljena je sustavna analiza kao i priprema podataka. U cilju rješavanja navedenog problema, u radu su kreirani regresijski i klasifikacijski modeli te je napravljena klaster analiza sa svrhom optimizacije rezultata klasifikacijskog modela. U zaključnom dijelu rada dane su i smjernice za daljnja istraživanja.

Abstract

The subject of this thesis is application of machine learning methods in line production of fire dampers and variable flow regulators for the assembly line. The use of machine learning and deep data analysis has become a standard in all aspects of production. Its use is to reveal hidden information and knowledge in the data which greatly contributes to the decision making process. The aim of this thesis is to find a mathematical model that will analyze the data according to the criterion of accuracy and contribute to the understanding of the line production assembly process. The first part of the paper describes the theory behind machine learning, the techniques for deep data analysis – cluster methods, regression and classification. In the practical part of the thesis, systematic analysis and data preparation were performed on the data from the business system In order to address this problem, regression and classification models were created. To optimize the classification model results a cluster analysis was used. Further research guidelines are noted in the conclusion.

Item Type:	Thesis (Master's thesis (Bologna))
Uncontrolled Keywords:	strojno učenje; dubinska analiza podataka; klasifikacija; regresija; klaster analiza
Keywords (Croatian):	machine learning; deep data analysis; classification; regression; cluster analysis
Subjects:	TECHNICAL SCIENCE TECHNICAL SCIENCE > Mechanical Engineering TECHNICAL SCIENCE > Mechanical Engineering > mechanical engineering design and drafting
Divisions:	700 Department of Industrial Engineering > 720 Chair of Production Control
Date Deposited:	29 Nov 2018 09:22
Last Modified:	25 Oct 2019 11:05
URI:	http://repozitorij.fsb.hr/id/eprint/8941

Actions (login required)

View Item