Využití predikčních modelů v analýze nezaměstnanosti

The purpose of this diploma thesis is compare selected prediction models. These models are used for analysis of the South Moravian Region unemployment and then compared with unemployment development in the Czech Republic. This thesis is separated to two parts: theoretical and practical. The theoretical part is used to describe selected prediction models, psychological effects of unemployment on human being and impact of unemployment on the society. The second part is focused on the analysis of the South Moravian Region. Prediction models such as trend lines, moving averages, ARIMA method and neural network are used and compared which each other. All predictions are based on data from the Czech Statistical Office and covers period between year 1993 and 2012. This data is measured with periodicity of three months