Show simple item record

Application of machine learning to the problem of missing data in the development of predictive models

dc.contributor.advisorKupusinac, Aleksandar
dc.contributor.advisorDoroslovački, Ksenija
dc.contributor.otherIvetić, Dragan
dc.contributor.otherProtić, Jelica
dc.contributor.otherStokić, Edita
dc.contributor.otherSladić, Goran
dc.contributor.otherKupusinac, Aleksandar
dc.contributor.otherDoroslovački, Ksenija
dc.creatorVrbaški, Dunja
dc.date.accessioned2020-08-28T15:14:55Z
dc.date.available2020-08-28T15:14:55Z
dc.date.issued2020-07-20
dc.identifier.urihttps://www.cris.uns.ac.rs/DownloadFileServlet/Disertacija158472080057127.pdf?controlNumber=(BISIS)114270&fileName=158472080057127.pdf&id=15107&source=NaRDuS&language=srsr
dc.identifier.urihttps://www.cris.uns.ac.rs/record.jsf?recordId=114270&source=NaRDuS&language=srsr
dc.identifier.urihttps://www.cris.uns.ac.rs/DownloadFileServlet/IzvestajKomisije158472081075787.pdf?controlNumber=(BISIS)114270&fileName=158472081075787.pdf&id=15108&source=NaRDuS&language=srsr
dc.identifier.uri/DownloadFileServlet/IzvestajKomisije158472081075787.pdf?controlNumber=(BISIS)114270&fileName=158472081075787.pdf&id=15108
dc.identifier.urihttp://nardus.mpn.gov.rs/handle/123456789/17378
dc.description.abstractProblem nedostajućih podataka je često prisutan prilikom razvoja prediktivnih modela. Umesto uklanjanja podataka koji sadrže vrednosti koje nedostaju mogu se primeniti metode za njihovu imputaciju. Disertacija predlaže metodologiju za pristup analizi uspešnosti imputacija prilikom razvoja prediktivnih modela. Na osnovu iznete metodologije prikazuju se rezultati primene algoritama mašinskog učenja, kao metoda imputacije, prilikom razvoja određenih, konkretnih prediktivnih modela.sr
dc.description.abstractThe problem of missing data is often present when developing predictive models. Instead of removing data containing missing values, methods for imputation can be applied. The dissertation proposes a methodology for analysis of imputation performance in the development of predictive models. Based on the proposed methodology, results of the application of machine learning algorithms, as an imputation method in the development of specific models, are presented.en
dc.languagesr (latin script)
dc.publisherУниверзитет у Новом Саду, Факултет техничких наукаsr
dc.rightsopenAccessen
dc.sourceУниверзитет у Новом Садуsr
dc.subjectmašinsko učenjesr
dc.subjectMachine learningen
dc.subjectnedostajući podacisr
dc.subjectprediktivni modelisr
dc.subjectveštačke neuralne mrežesr
dc.subjectslučajne šumesr
dc.subjectmissing dataen
dc.subjectpredictive modelingen
dc.subjectartificial neuralnetworksen
dc.subjectrandom forestsen
dc.titlePrimena mašinskog učenja u problemu nedostajućih podataka pri razvoju prediktivnih modelasr
dc.title.alternativeApplication of machine learning to the problem of missing data in the development of predictive modelsen
dc.typedoctoralThesisen
dc.rights.licenseBY-NC-ND
dc.identifier.fulltexthttp://nardus.mpn.gov.rs/bitstream/id/65051/IzvestajKomisije.pdf
dc.identifier.fulltexthttp://nardus.mpn.gov.rs/bitstream/id/65050/Disertacija.pdf


Files in this item

Thumbnail
Thumbnail

This item appears in the following Collection(s)

Show simple item record