prikaz prve stranice dokumenta Kolaborativno filtriranje
  Download
PDF 845.55 KB
master's thesis
Kolaborativno filtriranje
2016. urn:nbn:hr:126:290773
Marković, Mirna
Josip Juraj Strossmayer University of Osijek
Department of Mathematics
Chair of Applied Mathematics
Numerical Mathematics Research Group

Cite this document

Marković, M. (2016). Kolaborativno filtriranje (Master's thesis). Retrieved from https://urn.nsk.hr/urn:nbn:hr:126:290773

Marković, Mirna. "Kolaborativno filtriranje." Master's thesis, Josip Juraj Strossmayer University of Osijek, Department of Mathematics, 2016. https://urn.nsk.hr/urn:nbn:hr:126:290773

Marković, Mirna. "Kolaborativno filtriranje." Master's thesis, Josip Juraj Strossmayer University of Osijek, Department of Mathematics, 2016. https://urn.nsk.hr/urn:nbn:hr:126:290773

Marković, M. (2016). 'Kolaborativno filtriranje', Master's thesis, Josip Juraj Strossmayer University of Osijek, Department of Mathematics, accessed 17 May 2022, https://urn.nsk.hr/urn:nbn:hr:126:290773

Marković M. Kolaborativno filtriranje [Master's thesis]. Osijek: Josip Juraj Strossmayer University of Osijek, Department of Mathematics; 2016 [cited 2022 May 17] Available at: https://urn.nsk.hr/urn:nbn:hr:126:290773

M. Marković, "Kolaborativno filtriranje", Master's thesis, Josip Juraj Strossmayer University of Osijek, Department of Mathematics, Osijek, 2016. Available at: https://urn.nsk.hr/urn:nbn:hr:126:290773

Please login to the repository to save this object to your list.
Metadata
TitleKolaborativno filtriranje
Title (english)Collaborative Filtering
AuthorMirna Marković
MentorZoran Tomljanović (mentor)
Committee memberKristian Sabo (predsjednik povjerenstva)
Committee memberZoran Tomljanović (član povjerenstva)
Committee memberIvana Kuzmanović (član povjerenstva)
GranterJosip Juraj Strossmayer University of Osijek
Department of Mathematics
(Chair of Applied Mathematics)
(Numerical Mathematics Research Group)
Osijek
Defense date and country2016-12-16, Croatia
Scientific / art field,
discipline and subdiscipline		NATURAL SCIENCES
Mathematics
Abstract
U radu je opisan jedan tip sustava za preporuke, točnije kolaborativno filtriranje bazirano na modelu orijentiranom prema proizvodu i pripadni algoritam baziran na SVD dekompoziciji matrice sustava. Ukratko je dan kratak povijesni pregled sustava za preporuke te su navedeni osnovni pojmovi i opisani su najbitniji tipovi sustava za preporuke. Zatim je iskazan centralni algoritam rada, algoritam za preporuke zasnovan na SVD dekompoziciji s redukcijom dimenzije te je obrazložena njegova korektnost. Kroz svojstva navedenog algoritma naglašena je korisnost SVD dekompozicije matrica u stvarnom svijetu. Nakon toga opisana su dva načina za osvje- žavanje sustava novim korisnicima, folding-in metoda i update metoda te su dani pripadni algoritmi. U zadnjem dijelu rada opisani su testovi provedeni na implementiranim algoritmima, algoritmu za preporuke zasnovanom na SVD dekompoziciji s redukcijom dimenzije i folding-in metodi, a pripadni Matlab kodovi priloženi su u Prilogu 2 i 3.
Abstract (english)
The thesis is dealing with a special type of recommender system, with the main focus on the item and model based collaborative filtering recommender system including corresponding SVD based algorithm. The historical overview of recommender systems is briefly illustrated, key terms are listed and the most important types of recommender systems are described. Afterwards, the central algorithm of the thesis, the SVD algorithm for recommender system with dimensionality reduction, has been stated and its utility is motivated. The benefit of SVD decomposition in the real world has been outlined while describing properties of the central algorithm. After that, the folding-in method and the update method for updating system with new users have been described and related algorithms are given. The last section of the thesis displays the test examples of implemented algorithms, the central algorithm of the thesis and the folding-in method. Code implementation using Matlab is appended at the end of the thesis in Appendix 2 and Appendix 3.
Keywords
Keywords (english)
Languagecroatian
URN:NBNurn:nbn:hr:126:290773
Study programmeTitle: Mathematics; specializations in: Financial and Statistical Mathematics, Mathematics and Computer Science, Industrial and Applied Mathematics
Course: Financial and Statistical Mathematics
Study programme type: university
Study level: graduate
Academic / professional title: magistar/magistra matematike (magistar/magistra matematike)
Type of resourceText
File originBorn digital
Access conditionsOpen access
Terms of useLicence In copyright icon
Created on2016-12-19 07:35:07