o:1292 Doprinos analizi vremenskih nizova sa celobrojnim vrednostima Contributions to the analysis of integer-valued time series : doctoral dissertation sr New thinning operators are defined as differences of two negative binomial thinning operators. On the basis of such defined operators, time series with discrete Laplace marginals are defined. Some important features of all introduced models are determined. Estimators of unknown parameters are derived and their asimptotic behaviour are discussed. All models are checked on simulated data and compared with some of existing models. An application in real-life situations are presented. Also, a method for identification of latent components of the models are given Biobibliografski podaci: str. [215] Datum odbrane: 12.12.2016. statistics for stochastic processes 1025419241 91552101 4846 91552100 1025419241 2017-04-19T19:08:41.053Z 45 no 46 mentor Miodrag S. 1974 - Đorđević 2016 63 mentor Miroslav M. 1972 - Ristić 2016 63 član komisije Biljana 1954 - Popović 2016 63 član komisije Miomir 1949 - Stanković 2016 63 član komisije Aleksandar 1978 - Nastić 2016 vi, 213 str. 3588984 http://phaidrabg.bg.ac.rs/o:1292 no yes 4 2017-04-19T19:08:41.320Z 70 DLINAR, SDLINAR, CSDLINAR, INAR, marginalnadiskretna Laplasova raspodela, negativnibinomni tining operator, ocenjivanje parametara,identifikacija latentnih komponenti DLINAR, SDLINAR, CSDLINAR, INAR, discreteLaplace marginal distribution, negative binomialthinning operator, parameter estimation,identification of latent components 519.246/.8(043.3) 1738 18A07 18A0701 2016