prikaz prve stranice dokumenta Razvoj sustava za prilagodljivu ugradnju i izvođenje postupaka i modela strojnog učenja
  Download
PDF 6.96 MB
master's thesis
Razvoj sustava za prilagodljivu ugradnju i izvođenje postupaka i modela strojnog učenja
Osijek: Josip Juraj Strossmayer University of Osijek, Faculty of Electrical Engineering, Computer Science and Information Technology Osijek, 2019. urn:nbn:hr:200:190243
Ivić, Vedran
Josip Juraj Strossmayer University of Osijek
Faculty of Electrical Engineering, Computer Science and Information Technology Osijek
Department of Software Engineering
Chair of Programming Languages and Systems

Institutional repository: Faculty of Electrical Engineering, Computer Science and Information Technology Osijek

Cite this document

Ivić, V. (2019). Razvoj sustava za prilagodljivu ugradnju i izvođenje postupaka i modela strojnog učenja (Master's thesis). Osijek: Josip Juraj Strossmayer University of Osijek, Faculty of Electrical Engineering, Computer Science and Information Technology Osijek. Retrieved from https://urn.nsk.hr/urn:nbn:hr:200:190243

Ivić, Vedran. "Razvoj sustava za prilagodljivu ugradnju i izvođenje postupaka i modela strojnog učenja." Master's thesis, Josip Juraj Strossmayer University of Osijek, Faculty of Electrical Engineering, Computer Science and Information Technology Osijek, 2019. https://urn.nsk.hr/urn:nbn:hr:200:190243

Ivić, Vedran. "Razvoj sustava za prilagodljivu ugradnju i izvođenje postupaka i modela strojnog učenja." Master's thesis, Josip Juraj Strossmayer University of Osijek, Faculty of Electrical Engineering, Computer Science and Information Technology Osijek, 2019. https://urn.nsk.hr/urn:nbn:hr:200:190243

Ivić, V. (2019). 'Razvoj sustava za prilagodljivu ugradnju i izvođenje postupaka i modela strojnog učenja', Master's thesis, Josip Juraj Strossmayer University of Osijek, Faculty of Electrical Engineering, Computer Science and Information Technology Osijek, accessed 13 May 2024, https://urn.nsk.hr/urn:nbn:hr:200:190243

Ivić V. Razvoj sustava za prilagodljivu ugradnju i izvođenje postupaka i modela strojnog učenja [Master's thesis]. Osijek: Josip Juraj Strossmayer University of Osijek, Faculty of Electrical Engineering, Computer Science and Information Technology Osijek; 2019 [cited 2024 May 13] Available at: https://urn.nsk.hr/urn:nbn:hr:200:190243

V. Ivić, "Razvoj sustava za prilagodljivu ugradnju i izvođenje postupaka i modela strojnog učenja", Master's thesis, Josip Juraj Strossmayer University of Osijek, Faculty of Electrical Engineering, Computer Science and Information Technology Osijek, Osijek, 2019. Available at: https://urn.nsk.hr/urn:nbn:hr:200:190243

Metadata
TitleRazvoj sustava za prilagodljivu ugradnju i izvođenje postupaka i modela strojnog učenja
Title (english)Development of system for flexible implementation and execution of machine learning models and algorithms
AuthorVedran Ivić
MentorGoran Martinović (mentor)
Committee memberGoran Martinović (predsjednik povjerenstva)
Committee memberČaslav Livada (član povjerenstva)
Committee memberBruno Zorić (član povjerenstva)
GranterJosip Juraj Strossmayer University of Osijek
Faculty of Electrical Engineering, Computer Science and Information Technology Osijek
(Department of Software Engineering)
(Chair of Programming Languages and Systems)
Osijek
Defense date and country2019-10-09, Croatia
Scientific / art field,
discipline and subdiscipline		TECHNICAL SCIENCES
Computing
Artificial Intelligence
Abstract
Tema ovog diplomskog rada je razvoj sustava za prilagodljivu ugradnju i izvođenje postupaka i modela strojnog učenja za primjenu u području vizualne analize sentimenta, koji je potrebno i programski ostvariti. Opisani su zahtjevi na takav sustav te analizirani prikladni postupci i modeli strojnog učenja koji omogućuju detekciju i analizu sentimenta sa slike. Proučeno je i obrađeno područje umjetnih neuronskih mreža s naglaskom na konvolucijske neuronske mreže. Pobliže je opisano područje primjene odnosno analize sentimenta, uz prikaz već postojećih sustava za implementiranje strojnog učenja u tu svrhu. Odabran je odgovarajući algoritam te je objašnjen njegov princip rada i parametri. Zatim su opisane i razine te funkcionalne komponente arhitekture sustava te je objašnjena njihova međusobna komunikacija. Koristeći adekvatna okruženja, alate i tehnologije (TensorFlow, Keras, Python, Node.js i drugi), koji su također pobliže objašnjeni u radu, sustav je i programski razvijen. Ponajprije je izrađena baza podataka slika, na osnovu koje je odabranim algoritmom treniran i dobiven model za predviđanje sentimenta sa slike. Zatim je programski ostvaren modul za predviđanje kao web aplikacija koja omogućuje upotrebu sustava krajnjem korisniku. Konačno, sustav je testiran po pitanju integracije korisničkog sučelja i korištenja te funkcionalnosti sustava, uz analizu dobivenih rezultata i osvrt na cjelokupan sustav.
Abstract (english)
The topic of this paper is the development of system for flexible implementation and execution of machine learning algorithms and models for application in the field of visual sentiment analysis as well as the realization of the very system programmatically. System requirements have been described and suitable macine learning algorithms and models that enable image sentiment analysis have been analyzed. The field of artificial neural networks has been studied and processed with emphasis on convolutional neural networks. The field of application, i.e. sentiment analysis has been elaborated in more detail, while presenting the already existing systems for machine learning implementation for this very application. An appropriate algorithm has been selected and its operating principle and parameters explained. Then, the levels and functional components of the system architecture have been described while explaining ways of their communication. Using the suitable environments, tools and technologies (TensorFlow, Keras, Python, Node.js and others), that had also been explained in the paper, the system has been programmatically developed. First, an image database had been created, based on which the selected algorithm has trained and obtained the model for predicting sentiment from an image. Then, the prediction module has been implemented as a web application that enables the system to be used by the end user. Finally, the system has been tested for the integration of the user interface and use, as well as for the functionalities of the system, with a provided analysis of the results obtained and a review of the overall system.
Keywords
Keywords (english)
Languagecroatian
URN:NBNurn:nbn:hr:200:190243
Study programmeTitle: Graduate University Study Programme in Electrical Engineering; specializations in: Communications and Informatics, Power Engineering
Course: Communications and Informatics
Study programme type: university
Study level: graduate
Academic / professional title: magistar/magistra inženjer/inženjerka elektrotehnike (magistar/magistra inženjer/inženjerka elektrotehnike)
Type of resourceText
File originBorn digital
Access conditionsOpen access
Terms of useLicence In copyright icon
Created on2020-01-15 11:52:41