|
Zhurnal Vychislitel'noi Matematiki i Matematicheskoi Fiziki, 2010, Volume 50, Number 5, Pages 967–976
(Mi zvmmf4883)
|
|
|
|
Minimax feature selection problem for constructing a classifier using support vector machines
Yu. V. Goncharov Dorodnicyn Computing Center, Russian Academy of Sciences, ul. Vavilova 40, Moscow, 119333 Russia
Abstract:
A minimax feature selection problem for constructing a classifier using support vector machines is considered. Properties of the solutions of this problem are analyzed. An improvement of the saddle point search algorithm based on extending the bound for the step parameter is proposed. A new nondifferential optimization algorithm is developed that, together with the saddle point search algorithm, forms a hybrid feature selection algorithm. The efficiency of the algorithm for computing Dykstra’s projections as applied for the feature selection problem is experimentally estimated.
Key words:
feature selection problem, minimax problem, support vector machine, saddle point searching algorithm, subgradient algorithm.
Received: 28.09.2009 Revised: 23.12.2009
Citation:
Yu. V. Goncharov, “Minimax feature selection problem for constructing a classifier using support vector machines”, Zh. Vychisl. Mat. Mat. Fiz., 50:5 (2010), 967–976; Comput. Math. Math. Phys., 50:5 (2010), 917–925
Linking options:
https://www.mathnet.ru/eng/zvmmf4883 https://www.mathnet.ru/eng/zvmmf/v50/i5/p967
|
|