Time series analysis of uninsured vehicles accidents

Thesis is focused on a time series analysis of uninsured vehicles accidents in Czech republic between years 2003 and 2011. The objective is to create a time series model with appropriate statistical methods to detect and quantify seasonal patterns in given data. The goal is to find a season or period within a year where significantly different levels of accidents occures. Final model belongs to the category of dynamic regression, containing dummy variables, deterministic trend and ARMA component