WP. 32 ENGLISH ONLY UNITED NATIONS STATISTICAL COMMISSION and ECONOMIC COMMISSION FOR EUROPE CONFERENCE OF EUROPEAN STATISTICIANS EUROPEAN COMMISSION STATISTICAL OFFICE OF THE EUROPEAN COMMUNITIES (EUROSTAT)
نویسنده
چکیده
Controlled tabular adjustment is an SDL methodology based on a mixed integer linear programming model. We develop new hybrid heuristics and new metaheuristic learning approaches for solving this model, and examine their performance. Our new approaches are based on partitioning the problem into its discrete and continuous components, and first creating a hybrid that reduces the number of binary variables through a grouping procedure that combines an exact mathematical programming model with constructive heuristics. We then replace the MILP with an evolutionary scatter search approach that extends the method to large problems with over 9000 entries. Finally, we introduce a new metaheuristic learning method that significantly improves the quality of solutions obtained.
منابع مشابه
WP.14 ENGLISH ONLY UNITED NATIONS STATISTICAL COMMISSION and ECONOMIC COMMISSION FOR EUROPE CONFERENCE OF EUROPEAN STATISTICIANS EUROPEAN COMMISSION STATISTICAL OFFICE OF THE EUROPEAN COMMUNITIES (EUROSTAT)
متن کامل
WP. 9 ENGLISH ONLY UNITED NATIONS STATISTICAL COMMISSION and ECONOMIC COMMISSION FOR EUROPE CONFERENCE OF EUROPEAN STATISTICIANS EUROPEAN COMMISSION STATISTICAL OFFICE OF THE EUROPEAN COMMUNITIES (EUROSTAT)
The concept of differential privacy has received considerable attention in the literature recently. In this paper we evaluate the masking mechanism based on Laplace noise addition to satisfy differential privacy. The results of this study indicate that the Laplace based noise addition procedure does not satisfy the requirements of differential privacy.
متن کاملWP.2 ENGLISH ONLY UNITED NATIONS STATISTICAL COMMISSION and ECONOMIC COMMISSION FOR EUROPE CONFERENCE OF EUROPEAN STATISTICIANS EUROPEAN COMMISSION STATISTICAL OFFICE OF THE EUROPEAN COMMUNITIES (EUROSTAT)
The European Union is facing the problem of releasing microdata in a multinational setting i.e. microdata stemming from twenty seven member states. Different laws, methodologies, practices and cultural approaches to confidentiality may severely limit the possibility of obtaining comprehensive anonymised data sets. We recall the approach adopted in Europe to design and manage harmonised statisti...
متن کاملUNITED NATIONS STATISTICAL COMMISSION and ECONOMIC COMMISSION FOR EUROPE CONFERENCE OF EUROPEAN STATISTICIANS EUROPEAN COMMISSION STATISTICAL OFFICE OF THE EUROPEAN COMMUNITIES (EUROSTAT) ORGANIZATION FOR ECONOMIC COOPERATION AND DEVELOPMENT (OECD) STATISTICS DIRECTORATE
متن کامل
WP. 30 ENGLISH ONLY UNITED NATIONS STATISTICAL COMMISSION and ECONOMIC COMMISSION FOR EUROPE CONFERENCE OF EUROPEAN STATISTICIANS EUROPEAN COMMISSION STATISTICAL OFFICE OF THE EUROPEAN COMMUNITIES (EUROSTAT)
We extend the safety rules used for the Statistical Disclosure Control of magnitude tables to include an intruder who models the ignorance about an unknown confidential quantity with a Uniform distribution. By applying this extension to the generalised p-rule we obtain the safety rules useful also in the presence of groups of respondents. The corresponding disclosure rules for different prior k...
متن کاملWP. 33 ENGLISH ONLY UNITED NATIONS STATISTICAL COMMISSION and ECONOMIC COMMISSION FOR EUROPE CONFERENCE OF EUROPEAN STATISTICIANS EUROPEAN COMMISSION STATISTICAL OFFICE OF THE EUROPEAN COMMUNITIES (EUROSTAT)
In order to manage the disclosure risk in frequency tables containing population counts, the tables undergo statistical disclosure control (SDC) methods. This results in information loss. We examine quantitative information loss measures for frequency tables and compare them across different SDC methods. We show examples of the information loss measures on real UK 2001 Census tables after they ...
متن کامل