prikaz prve stranice dokumenta Priprema skupova bioloških podataka za strojno i duboko učenje
No public access
undergraduate thesis
Priprema skupova bioloških podataka za strojno i duboko učenje
2024. urn:nbn:hr:181:123884
Dragan, David
Josip Juraj Strossmayer University of Osijek
Department of biology

Cite this document

Dragan, D. (2024). Priprema skupova bioloških podataka za strojno i duboko učenje (Undergraduate thesis). Osijek: Josip Juraj Strossmayer University of Osijek, Department of biology. Retrieved from https://urn.nsk.hr/urn:nbn:hr:181:123884

Dragan, David. "Priprema skupova bioloških podataka za strojno i duboko učenje." Undergraduate thesis, Josip Juraj Strossmayer University of Osijek, Department of biology, 2024. https://urn.nsk.hr/urn:nbn:hr:181:123884

Dragan, David. "Priprema skupova bioloških podataka za strojno i duboko učenje." Undergraduate thesis, Josip Juraj Strossmayer University of Osijek, Department of biology, 2024. https://urn.nsk.hr/urn:nbn:hr:181:123884

Dragan, D. (2024). 'Priprema skupova bioloških podataka za strojno i duboko učenje', Undergraduate thesis, Josip Juraj Strossmayer University of Osijek, Department of biology, accessed 22 October 2024, https://urn.nsk.hr/urn:nbn:hr:181:123884

Dragan D. Priprema skupova bioloških podataka za strojno i duboko učenje [Undergraduate thesis]. Osijek: Josip Juraj Strossmayer University of Osijek, Department of biology; 2024 [cited 2024 October 22] Available at: https://urn.nsk.hr/urn:nbn:hr:181:123884

D. Dragan, "Priprema skupova bioloških podataka za strojno i duboko učenje", Undergraduate thesis, Josip Juraj Strossmayer University of Osijek, Department of biology, Osijek, 2024. Available at: https://urn.nsk.hr/urn:nbn:hr:181:123884

Please login to the repository to save this object to your list.
Metadata
TitlePriprema skupova bioloških podataka za strojno i duboko učenje
Title (english)Preparing biological data sets for machine and deep learning
AuthorDavid Dragan
MentorBranimir Hackenberger Kutuzović (mentor)
Committee memberBranimir Hackenberger Kutuzović (predsjednik povjerenstva)
GranterJosip Juraj Strossmayer University of Osijek
Department of biology
Osijek
Defense date and country2024-09-09, Croatia
Scientific / art field,
discipline and subdiscipline		NATURAL SCIENCES
Biology
Abstract
U radu su protumačeni osnovni pojmovi vezani uz područje strojnog i dubokog učenja, ukratko je opisan proces klasifikacije zemljišnog pokrova te metode klasifikacije zemljišnog pokrova. Cilj rada bio je klasificirati zemljišni pokrov jednog dijela područja južno od grada Osijeka u programu QGIS te usporediti preciznost i efikasnost algoritama Minimum Distance i Spectral Angle Mapping u sklopu dodatka za QGIS zvanog Semi-Automatic Classification Plugin. Iz rezultata vidljivo je kako je algoritam Minimum Distance bio efikasniji i precizniji od algoritma Spectral Angle Mapping, iako nijedan od algoritama nije prikazao dovoljnu preciznost da bi se obrađeni raster mogao smatrati dovoljno preciznom preslikom stvarnog stanja na terenu. Na kraju su ponuđeni mogući razlozi pojave grešaka i nepreciznosti u obradi rastera te moguća rješenja koja bi ubuduće mogla pomoći u kvalitetnijoj obradi istog rastera
Abstract (english)
In this work, concepts of machine and deep learning are explained, and in short, process of land cover classification and land cover classification methods are described. The goal of this work was to classify land cover of one part of the land located southward of the city of Osijek. The classification process was carried out in program called QGIS. Beside the classification of the land, the goal was also to compare precision and efficiency of algorithms called Minimum Distance and Spectral Angle Mapping, which are part of the add-on for QGIS called Semi-Automatic Classification Plugin. From the results it is clear that algorithm Minimum Distance performed better then Spectral Angle Mapping, although neither one of them provided enough precision that we could say that the classified map showed actual situation on the field. At the end, author offered possible reasons that could have caused occurrence of mistakes in the classified map and the possible solutions for resolution of the problem.
Keywords
Keywords (english)
Languagecroatian
URN:NBNurn:nbn:hr:181:123884
Study programmeTitle: Undergraduate University Study Programme in Biology (Bachelor of Science)
Study programme type: university
Study level: undergraduate
Academic / professional title: sveučilišni/a prvostupnik/prvostupnica (baccalaureus/baccalaurea) biologije (sveučilišni/a prvostupnik/prvostupnica (baccalaureus/baccalaurea) biologije)
Type of resourceText
File originBorn digital
Access conditionsClosed access
Terms of useLicence In copyright icon
Created on2024-10-17 08:27:31