Institutional Repository of Vilnius University
Bendrųjų namų ūkio pajamų ir socialinių išmokų analizė /
Abstract
dc:descriptionThis paper analyzes annual household income in Lithuania, based on data from 2023 collected through the “Statistics on Income and Living Conditions” survey, coordinated by the State Data Agency. The aim of the study is to comprehensively evaluate total household income and the significance of social benefits by applying various statistical and advanced analytical methods. The analysis is focused on understanding how income is distributed among different population groups and which factors have the greatest influence on income levels. The study begins with an initial data analysis -- evaluating income distribution, statistical indicators, and identifying possible irregularities in the data. This initial exploration provided a strong foundation for selecting appropriate models to fulfill the study’s objective. A linear regression model was tested first and provided some useful insights, but it ultimately failed to meet the study goals due to unsatisfied assumptions. The quantile regression model did not yield satisfactory results either, which led to the application of a more flexible model -- GAMLSS. The GAMLSS model was implemented with several different distributions to ensure the best possible fit to the data. Given the often heterogeneous nature of income distribution, the dataset was further divided into subgroups -- households with lower and higher income levels. This allowed for a more detailed evaluation of model performance in different segments and helped to identify structural differences that may not have been visible in the overall data. This segmentation provided a solid foundation for further analysis and contributed to improved results. In addition to traditional statistical models, three machine learning models were applied. These models helped assess how accurately income could be predicted based on selected variables and allowed a comparison of their predictive performance with classical statistical approaches. The use of different models produced favorable results -- it was possible to construct models well-suited to the data at hand. In summary, the study's findings are relevant for both social policy development and economic analysis, offering deeper insights into income inequality and the role of social benefits across different societal groups.
Degree
thesis:*- Grantor dc:publisher
- Institutional Repository of Vilnius University
- Year dc:date
- 2025
Author and committee
dc:creator, dc:contributor.*- Author dc:creator
-
- Astrauskas, Rokas,
Rights
dc:rights- Statement dc:rights
-
- info:eu-repo/semantics/openAccess
- Language dc:language
- lit
Identifiers
dc:identifier.*- Repository record dc:identifier
- https://repository.vu.lt/VU:ELABAETD239028343&prefLang=en_US