prikaz prve stranice dokumenta Model za segmentiranje kupaca pomoću metoda strojnog učenja
Access restricted to students and staff of home institution
master's thesis
Model za segmentiranje kupaca pomoću metoda strojnog učenja
2019. urn:nbn:hr:145:606615
Iličić, Mato
Josip Juraj Strossmayer University of Osijek
Faculty of Economics and Business in Osijek
Chair of Quantitative Methods and Informatics

Cite this document

Iličić, M. (2019). Model za segmentiranje kupaca pomoću metoda strojnog učenja (Master's thesis). Osijek: Josip Juraj Strossmayer University of Osijek, Faculty of Economics and Business in Osijek. Retrieved from https://urn.nsk.hr/urn:nbn:hr:145:606615

Iličić, Mato. "Model za segmentiranje kupaca pomoću metoda strojnog učenja." Master's thesis, Josip Juraj Strossmayer University of Osijek, Faculty of Economics and Business in Osijek, 2019. https://urn.nsk.hr/urn:nbn:hr:145:606615

Iličić, Mato. "Model za segmentiranje kupaca pomoću metoda strojnog učenja." Master's thesis, Josip Juraj Strossmayer University of Osijek, Faculty of Economics and Business in Osijek, 2019. https://urn.nsk.hr/urn:nbn:hr:145:606615

Iličić, M. (2019). 'Model za segmentiranje kupaca pomoću metoda strojnog učenja', Master's thesis, Josip Juraj Strossmayer University of Osijek, Faculty of Economics and Business in Osijek, accessed 12 January 2025, https://urn.nsk.hr/urn:nbn:hr:145:606615

Iličić M. Model za segmentiranje kupaca pomoću metoda strojnog učenja [Master's thesis]. Osijek: Josip Juraj Strossmayer University of Osijek, Faculty of Economics and Business in Osijek; 2019 [cited 2025 January 12] Available at: https://urn.nsk.hr/urn:nbn:hr:145:606615

M. Iličić, "Model za segmentiranje kupaca pomoću metoda strojnog učenja", Master's thesis, Josip Juraj Strossmayer University of Osijek, Faculty of Economics and Business in Osijek, Osijek, 2019. Available at: https://urn.nsk.hr/urn:nbn:hr:145:606615

Please login to the repository to save this object to your list.
Metadata
TitleModel za segmentiranje kupaca pomoću metoda strojnog učenja
Title (english)Model for client segmentation with machine learning methods
AuthorMato Iličić
MentorMarijana Zekić Sušac (mentor)
Committee memberJosip Mesarić (predsjednik povjerenstva)
Committee memberMartina Briš Alić (član povjerenstva)
Committee memberMarijana Zekić Sušac (član povjerenstva)
GranterJosip Juraj Strossmayer University of Osijek
Faculty of Economics and Business in Osijek
(Chair of Quantitative Methods and Informatics)
Osijek
Defense date and country2019-07-01, Croatia
Scientific / art field,
discipline and subdiscipline		SOCIAL SCIENCES
Economics
Business Informatics
Abstract
Ovaj rad bavi se klasifikacijom klijenata pomoću metoda strojnog učenja kako bi se identificirali klijenti koji će prihvatiti ponuđenu uslugu oročene štednje. Klijenti se klasificiraju u dvije skupine: potencijalni klijenti koji prihvaćaju i oni koji odbijaju ponuđenu uslugu oročene štednje. Podaci pomoću kojih se u radu kreira model neuronskih mreža i model stabla odlučivanja prikupljeni su pomoću telemarketinga u razdoblju od 2008. do 2013. godine od strane portugalskih banaka. Na kraju marketinških kampanja postotak pristanaka na ponuđenu oročenu štednju iznosio je 11,27%, što za banke nije bilo dovoljno učinkovito. Za problem klasifikacije klijenata kreiran je model pomoću metoda stabala odlučivanja i neuronskih mreža. Kako bi se povećala stopa prihvaćanja ponuđene usluge oročene štednje u budućim kampanjama, u radu je kreirano više modela neuronskih mreža i stabala odlučivanja kako bi se na kraju mogao izabrati najuspješniji model. Najvažniji pokazatelj uspješnosti modela je ukupna stopa točnosti klasifikacije na uzorku za validaciju. Uz prikaze rezultata svih modela, prikazat će se matrica konfuzije najboljih mreža i stabala odlučivanja te njihove analize osjetljivosti. Dva najbolja modela međusobno će se uspoređivati da bi se mogao odabrati najbolji model koji se preporučuje za uporabu uz smjernice za uporabu modela i za daljnja istraživanja.
Abstract (english)
This paper deals with the classification of clients by using machine learning methods in order to identify clients who would accept a term deposit service. The clients are classified in two groups: potential clients who accept and those who refuse the offered term deposit service. The information used to create a model of neural networks and the decision tree model were collected by telemarketing in the period between 2008 and 2013 by Portuguese banks. At the end of the marketing campaigns, the percentage of acceptations of the offered term deposit was 11.27%, which was not efficient enough for the banks. The decision tree model used in the paper was used for the classification problem while the neural networks model, in addition to the classification problem, also solved the forecasting problem. In order to increase the acceptance rate of the offered term deposit service in future campaigns, several neural network and decision tree models were created to finally choose the most successful model. The most important indicator of the success of the model was the overall classification accuracy rate on the validation sample. In addition to the results of all models, a confusion matrix of the best networks, decision trees, and their sensitivity analysis will be shown. The two best models will be compared in order to choose the best model recommended for use with the guidelines for the use of the model and for further research.
Keywords
Keywords (english)
Languagecroatian
URN:NBNurn:nbn:hr:145:606615
Study programmeTitle: Business economy; specializations in: Business Informatics
Course: Business Informatics
Study programme type: university
Study level: graduate
Academic / professional title: Magistar/magistra ekonomije (Magistar/magistra ekonomije)
Type of resourceText
File originBorn digital
Access conditionsAccess restricted to students and staff of home institution
Terms of useLicence In copyright icon
Created on2020-02-11 15:17:49