Abstract:
The problem of linear binary separation of finite interval sets (classes) is considered. Using the theory of interval systems of linear inequalities, the problem is reduced to the problem of finding a solution to a system of linear inequalities of a special form. In turn, the problem of finding a quasioptimal solution to the specified system (or a pseudo-solution in the case of its incompatibility and linear inseparability of classes) is reduced to problems of unconditional minimization. Illustrative numerical examples are given.
Keywords:classification, machine learning, interval data uncertainty.