| Title | Modelling bivariate ineger valued time series with seasonality: evidence on car accident data / |
|---|---|
| Translation of Title | Dvimačių sveikareikšmių sezoniškų laiko eilučių modeliai: autovarijų skaičiaus atvejis. |
| Authors | Rinkevičiūtė, Aušrinė |
| Full Text |
|
| Pages | 46 |
| Keywords [eng] | Sezoniškumas, dvimatis INAR, dvimatis INGARCH, laike kintantis BINAR, dvimatis Puasono skirstinys, avarijų skaičius. Seasonality, Bivariate INAR, Bivariate INGARCH, Time Variant BINAR, bivariate Poisson distribution, number of car accidents |
| Abstract [eng] | Although models for time series of counts are currently intensively studied by a number of researchers, analysis on how integer-valued time series with exhibited seasonality can be modelled is still limited. This thesis is focused on formulating models for bivariate integer-valued time series with exhibited seasonality. Hence, three models suitable to such are formulated: BINAR(1), BINGARCH and TV-BINAR(1). The considered estimation methods are tested on the simulated data. All three models are applied on the bivariate car accident data. Time series consist of number of car accidents caused by the alcohol intoxicated drivers and a number of car accidents caused by the sober drivers in Lithuania per month. |
| Dissertation Institution | Vilniaus universitetas. |
| Type | Master thesis |
| Language | English |
| Publication date | 2020 |