Modely statistické ochrany důvěrnosti mikrodat v populačních censech

In today's information environment, there is an increasing emphasis on data-driven decision making and ensuring access to relevant data sets. The census is a key source of information in official statistics, and the dissemination of microdata is becoming an important topic in the field of statistical confidentiality, motivated by advances in data analysis methods, dataset linkage, and information mining from data. Securing the confidentiality of microdata is an important aspect, as statistical offices are legally obliged to protect the information they collect. The aim of this dissertation is to describe in detail the issue of statistical confidentiality protection of microdata and to propose a practically feasible solution for dissemination of microdata from population censuses in the Czech Republic. In order to achieve these goals, in-house research was conducted, which included the development and preparation of a solution in the R programming language and active international collaboration. The process was supported by a thorough study of the relevant literature and analysis of the works of international experts. This dissertation makes a scientific contribution by addressing a topic that has not yet been studied in detail in the Czech Republic and providing practical approaches for its solution. Another contribution is the dissemination of microdata from population censuses for the Czech Republic, which represents a new and previously unavailable source of data for the scientific community. The paper also presents improvements in methods for simulating synthetic microdata and controlling the rate of information loss. Simulation of synthetic microdata has been identified as the most effective method of protecting population census microdata. This approach allows the creation of a new data set with maximum protection and minimum information loss while preserving the structure and relationships of the original data set. The obtained results have a high potential for practical application, especially at the Czech Statistical Office, where they can significantly facilitate the dissemination of microdata and datasets of official statistics for the scientific community.