prikaz prve stranice dokumenta Neuralne mreže u astrofizici
Access restricted to students and staff of home institution
undergraduate thesis
Neuralne mreže u astrofizici
2019. urn:nbn:hr:194:878697
Mrakovčić, Karlo
University of Rijeka
Faculty of Physics

Cite this document

Mrakovčić, K. (2019). Neuralne mreže u astrofizici (Undergraduate thesis). Rijeka: University of Rijeka. Retrieved from https://urn.nsk.hr/urn:nbn:hr:194:878697

Mrakovčić, Karlo. "Neuralne mreže u astrofizici." Undergraduate thesis, University of Rijeka, 2019. https://urn.nsk.hr/urn:nbn:hr:194:878697

Mrakovčić, Karlo. "Neuralne mreže u astrofizici." Undergraduate thesis, University of Rijeka, 2019. https://urn.nsk.hr/urn:nbn:hr:194:878697

Mrakovčić, K. (2019). 'Neuralne mreže u astrofizici', Undergraduate thesis, University of Rijeka, accessed 08 May 2024, https://urn.nsk.hr/urn:nbn:hr:194:878697

Mrakovčić K. Neuralne mreže u astrofizici [Undergraduate thesis]. Rijeka: University of Rijeka; 2019 [cited 2024 May 08] Available at: https://urn.nsk.hr/urn:nbn:hr:194:878697

K. Mrakovčić, "Neuralne mreže u astrofizici", Undergraduate thesis, University of Rijeka, Rijeka, 2019. Available at: https://urn.nsk.hr/urn:nbn:hr:194:878697

Please login to the repository to save this object to your list.
Metadata
TitleNeuralne mreže u astrofizici
Title (english)Neural networks in astrophysics
AuthorKarlo Mrakovčić
MentorDijana Dominis Prester (mentor)
Committee memberTomislav Terzić (predsjednik povjerenstva)
Committee memberDijana Dominis Prester (član povjerenstva)
Committee memberIvana Poljančić Beljan (član povjerenstva)
GranterUniversity of Rijeka
(Faculty of Physics)
Rijeka
Defense date and country2019-09-02, Croatia
Scientific / art field,
discipline and subdiscipline		NATURAL SCIENCES
Physics
Abstract
Neuralne mreže se sve vise koriste u računalnoj tehnologiji, te zbog svoje agilnosti u rješavanju širokog spektra problema, nalaze primjenu u svim sferama ljudskog djelovanja. Cilj ovog rada je predstaviti neuralnu mrežu, na najjednostavniji način objasniti njezinu strukturu, te prikazati potencijal neuralnih mreža u rješavanju složenih računalnih problema. U radu su razmatrane primjene neuralnih mreža u znanosti, s naglaskom na obradu podataka iz astrozike, gdje je primjena neuralnih mreža vrlo učinkovita. Uz povijesni pregled objašnjene su različite vrste strojnog učenja i način rada neuralnih mreža. Koristeći simulirane podatke teleskopa MAGIC, i samostalno izrađeni računalni program neuralne mreže ZvijezdANNa u programskom jeziku Python, napravljena je klasikacija čestica koje stvaraju atmosferske pljuskove na hadrone i fotone, rezultirajući sa 78% točnosti.
Abstract (english)
Neural networks are increasingly used in computer technology. Due to their eciency in solving a wide range of problems, they are nding applications in all spheres of human activities. The aim of this work is to present an example of a neural network, to explain its structure in a simple way, and to demonstrate its potential in solving complex computational problems. Ap- plications of neural networks in science are discussed, with an emphasis on analysing astrophysical data, where usage of neural networks is very eci- ent. A historical overview, and an explanation of dierent types of machine learning, including neural networks, is discussed. Using simulated data from the MAGIC telescopes and an independently developed neural network code ZvijezdANNa in Python programming language, a classication of cosmic ray air showers into hadron origin and photon origin has been performed, resulting in 78% accuracy.
Keywords
Languagecroatian
URN:NBNurn:nbn:hr:194:878697
Study programmeTitle: bacc. physics
Study programme type: university
Study level: undergraduate
Academic / professional title: sveučilišni/a prvostupnik/prvostupnica (baccalaureus/baccalaurea) fizike (sveučilišni/a prvostupnik/prvostupnica (baccalaureus/baccalaurea) fizike)
Type of resourceText
File originBorn digital
Access conditionsAccess restricted to students and staff of home institution
Terms of useLicence In copyright icon
Created on2019-11-30 13:34:22