350 rub
Journal Biomedical Radioelectronics №8 for 2016 г.
Article in number:
Fischer-s linear discriminant analysis in problems of multidimensional biomedical data classification
Authors:
A.P. NemirkoDr.Sc. (Eng.), Professor, Department of Biotechnical Systems, Saint Petersburg Electrotechnical University «LETI»E-mail: apn-bs@yandex.ru
Abstract:
For reduction of feature space dimension at biomedical data classification the Fischer-s linear discriminant is used. It reduces space dimension from initial to one by projecting the multidimensional data into a straight line. Pilot studies show that Fischer's criterion is not always optimum for the solution of an object recognition problem. Introduction of an additional weight vector can reduce an intersection between classes and lead to more effective procedures of linear classification on the plane. Recurrent expression for consecutive calculation of additional features is derived. When only one additional weight vector is used, procedure of classification is realized on the plane. Such approach demands introduction of other ways of class proximity measurements different from Fischer's criterion. In this work some ways of an assessment of class adjacency and degree of their intersection are offered. All these ways are applied to classification in space of two weight vectors found by Fischer's criterion.
Pages: 4-8
References

 

  1. Nemirko A.P., Manilo L.A., Kalinichenko A.N. Matematicheskie metody analiza biomedicinskikh dannykh. SPb: Izd-vo SPbGEHTU «LEHTI». 2013. 175 s.
  2. Rangajjjan R.M. Analiz biomedicinskikh signalov. Prakticheskijj podkhod / Per. s angl. A.N. Kalinichenko / Pod red. A.P. Nemirko. M.: Fizmatlit. 2007. 440 s.
  3. Ajjvazjan S.A., Bezhaeva Z.I., Staroverov O.V. Klassifikacija mnogomernykh nabljudenijj. M.: Statistika. 1974. 240 s.
  4. Duda R., KHart P. Raspoznavanie obrazov i analiz scen / Per. s angl. M.: Mir. 1976. 511 s.
  5. Nemirko A.P. Transformation of feature space based on Fisher-s linear discriminant. // Pattern Recognition and Image Analysis. Advances in Mathematical Theory and Applications. 2016. V. 26. № 2. R. 257-261.
  6. Manilo L.A. Uporjadochenie spektralnykh priznakov po ehmpiricheskim ocenkam mezhgruppovogo rasstojanija v zadachakh klassifikacii biosignalov // Izvestija vuzov Rossii. Radioehlektronika. 2006. Vyp. 3. S. 20-29.
  7. Manilo L.A. Linejjnyjj diskriminant Fishera v zadachakh raspoznavanija biosignalov po chastotnym svojjstvam // Sb. dokl. 12-jj Vseros. konf. «Matematicheskie metody raspoznavanija obrazov» (Moskva, 2005). M.: MAKS Press. 2005. S. 371-374.
  8. Gilbert E.G., Johnson D.W., Keerthi S.S. A fast procedure for computing the distance between complex objects in three-dimensional space // IEEE Journal of Robotics and Automation. April 1988. V. 4. R. 193-203.
  9. Iris Data Set. UCI Machine Learning Repository. https://archive.ics.uci.edu/ml/datasets/Iris, 2016.