(21) * << * >> * Russian * English * Content * All Issues

Algorithms of feature space formation by the criterion of conjugacy of measurement vectors

V.A. Fursov, V.A. Shustov

Image Processing Systems Institute of RAS, Samara

 PDF, 335 kB

Pages: 176-178

Full text of article: Russian language.

Abstract:
One of the central tasks of the general problem of pattern recognition is the issue of forming a feature space. Unfortunately, completely formalized methods for its solution are not yet available. Usually, two types of problems are considered within the framework of this general problem: enlarging the existing set of features with the new ones that improve the recognition quality and, conversely, excluding the least informative features from an existing redundant set (feature space reduction).

Citation:
Fursov VA, Shustov VA. Algorithms of feature space formation by the criterion of conjugacy of measurement vectors. Computer Optics 2001; 21: 176 - 178.

References:
  1. Fursov VA. Projections onto null-space in pattern recognition with a small number of observations [In Russian]. Proc All-Russian Conf on Pattern Recognition. Moscow: 1999: 119-121.
  2. Fursov VA. Training in Pattern Recognition from a small number of observations using projections onto null-space. Proc 15th Int Conf on Pattern recognition (ISPR) 2000; 2: 789-792.
  3. Fursov VA, Shustov VA. Formation of feature space by the criterion of conjugacy of measurement vectors. Computer Optics 2000; 20: 140-142.
  4. Duda RO, Hart PE. Pattern classification and scene analysis [In Russian]. Moscow: "Mir" Publisher; 1976.

© 2009, IPSI RAS
151, Molodogvardeiskaya str., Samara, 443001, Russia; E-mail: journal@computeroptics.ru ; Tel: +7 (846) 242-41-24 (Executive secretary), +7 (846) 332-56-22 (Issuing editor), Fax: +7 (846) 332-56-20