diplomsko delo
Mitja Žalik (Author), Niko Lukač (Mentor), Štefan Kohek (Co-mentor)

Abstract

V diplomskem delu predstavimo uporabo konvolucijskih nevronskih mrež za napoved geoprostorskih rastrskih podatkov. V prvem delu opišemo geoprostorske podatke in zgradbo ter značilnosti konvolucijskih nevronskih mrež. V drugem delu predlagamo model nevronske mreže, ki ga uporabimo za dolgoročno napoved sončnega potenciala in kratkoročno napoved vegetacijskega indeksa NDVI. Povprečna napaka po metriki NRMSE znaša 0,22% pri napovedi sončnega potenciala in 15% pri napovedi indeksa NDVI. Diplomsko delo zaključimo s predlogi možnih razširitev.

Keywords

umetna inteligenca;globoko učenje;konvolucijske nevronske mreže;geoprostorski podatki;rastrski podatki;diplomske naloge;

Data

Language: Slovenian
Year of publishing:
Typology: 2.11 - Undergraduate Thesis
Organization: UM FERI - Faculty of Electrical Engineering and Computer Science
Publisher: [M. Žalik]
UDC: 004.85:004.925(043.2)
COBISS: 40918019 Link will open in a new window
Views: 535
Downloads: 155
Average score: 0 (0 votes)
Metadata: JSON JSON-RDF JSON-LD TURTLE N-TRIPLES XML RDFA MICRODATA DC-XML DC-RDF RDF

Other data

Secondary language: English
Secondary title: Prediction of geospatial raster data using convolutional neural networks
Secondary abstract: In this thesis, the use of convolutional neural networks for predicting geospatial raster data is presented. In the first part, geospatial data are described. Then, the structure and characteristics of convolutional neural networks are explained. In the second part, we propose neural network model. It is used for long-term prediction of a solar potential and short-term prediction of normalized difference vegetation index (NDVI). The results are then evaluated. The Solar potential and NDVI index are predicted with average error 0.22% and 15% respectively, according to the NRMSE metric. The thesis is concluded with some suggestions for further enhancements.
Secondary keywords: artificial intelligence;deep learning;convolutional neural network;geospatial data;raster data;
Type (COBISS): Bachelor thesis/paper
Thesis comment: Univ. v Mariboru, Fak. za elektrotehniko, računalništvo in informatiko, Računalništvo in informacijske tehnologije
Pages: XIII, 37 f.
ID: 11973890