Mathematical methods for pattern recognition: logic, optimization, algebraic approaches
Vladimir Riazanov, Oleg Valentinovich Sen'ko, Yu. I. Zhuravlev · 2002
The paper is devoted to a brief survey of some new approaches and mathematical methods for solving recognition (supervised classification) problems developed by authors. The methods based on logical approach are first discussed. Next, multiparametric recognition model optimization (training) problems are studied. Finally, a construction of optimal solutions for recognition problems based on a two-level information handling process is presented.