o:1687 Nenegativni celobrojni autoregresivni procesi u slučajnoj sredini generisani geometrijskim brojačkim nizovima Random environment nonnegative integer-valued autoregressive processes generated by geometric counting series: doctoral dissertation : doctoral dissertation sr Here are analyzed integer autoregressive (INAR) processes in the random environment generated by geometric counting series. Firstly, the first order random environment INAR model is introduced. Later, random environment INAR models of higher order, as well as their general form, are defined. Finally, the bivariate model based on the bivariate random process is defined. The properties of all introduced models are analyzed. Estimation of unknown parameters is given and validate on the simulated data. Model quality is confirmed by application on the real-life data, comparing results with the competitive models. Biobibliografija: str. 113-115;Bibliografija: str. 109-112. Datum odbrane: 17.10.2020. Mathematical statistics, statistics of random processes 28177929 91552101 7995 91552100 28177929 2021-03-15T17:42:47.053Z 45 no 46 mentor Petra N. 1991- Laketa 2020 63 mentor Aleksandar 1978- Nastić 2020 63 predsednik komisije Miodrag Đorđević 2020 63 član komisije Miroslav Ristić 2020 63 član komisije Predrag Popović 2020 [8], 115 str. 2548869 http://phaidrabg.bg.ac.rs/o:1687 no yes 4 2021-03-15T17:42:47.320Z 70 INAR, NGINAR, slučajna sredina, negativni binomni tining operator, geometrijska marginalna raspodela, lanac Markova INAR, NGINAR, random environment, negative binomial thinning operator, geometric marginal distribution, Markov chain 519.246/.8(043.3) P160 1738 18A07 18A0701 2020