Institute of Mathematics with Computer Center at the Bulgarian Academy of Sciences
We study the two-group classification problem which involves classifying an observation into one of two groups based on its attributes. The classification rule is a hyperplane which separates the groups optimally,i.e. the set of the points of group 1 (2) belonging to the halfspace allocated for the points of group 2 (1) is of minimal cardinality. Exact and heuristic algorithms for solving the problem are presented. Computational results are reported.