Cell suppression problem: A genetic-based approach

Statistical confidentiality: Optimization techniques to protect tables

“…DeSilets et al [19] present a heuristic approach based on Neural Networks. Almeida et al [20] combine these ingredients in a genetic-based approach for protecting twodimensional tables with up to 100 rows and 100 columns. Chu [21] proposes a heuristic algorithm based on suppressing marginal cells.…”

A genetic approach to statistical disclosure control

“…However, the current tools leave much to be desired. As currently implemented, the output from the LP heuristic is not available to the user, and because of the large numbers of constraints and variables, the "optimal" approach is , and recently a hybrid genetic approach has been proposed to extend the scalability of this approach [1]. However these are not applicable to multi-dimensional or hierarchical tables.…”

A Genetic Approach to Statistical Disclosure Control

“…To give an example, an analysis of industrial activity broken down by region and activity type might have millions of cells and several dimensions, each with different levels of hierarchy. For two-dimensional non-hierarchical tables optimal methods based on a "network flow" formulation are possible for larger tables [10] [4], and recently a hybrid genetic approach has been proposed to extend the scalability of this approach [1]. However these are not applicable to multi-dimensional or hierarchical tables.…”