Multivariate Scaling of the Characteristic Features Based on Pseudo-Inverse Operations for Recognition Problems Solving

Abstract Some approach to multidimensional information scaling of the characteristics features based on results of theory of perturbation of pseudo-inverse and projective matrices and solutions of systems of linear algebraic equations is proposed. The method and algorithm of a piecewise hyperplane clusters creating with the verification of a given criterion for effectiveness of the proposed method of clustering is developed. The problem of stability of main indicators of the classifier in presence disturbances in source information is investigated. The proposed method for determining influence of source data errors on main indicators of the classifier provides presentation of undisturbed information matrix in form of splitting matrices of special kind. Advantages of the proposed approach are demonstrated, an example of using the method of scaling characteristic features for recognizing fingerspelling alphabet of sign language is given.

[1]  R. E. Cline.  Representations for the Generalized Inverse of a Partitioned Matrix , 1964 .

[2]  N. F. Kirichenko,et al.  Pseudoinverse and Projection Matrices in Problems of Synthesis of Functional Transformers , 2004 .

[3]  Methods for obtaining reliable solutions to systems of linear algebraic equations , 2011 .

[4]  Miguel Figueroa,et al.  Competitive learning with floating-gate circuits , 2002, IEEE Trans. Neural Networks.

[5]  Vladimir Vapnik,et al.  Statistical learning theory , 1998 .

[6]  Y. Takane,et al.  Generalized Inverse Matrices , 2011 .

[7]  Yuriy V. Krak,et al.  Recognition of Dactylemes of Ukrainian Sign Language Based on the Geometric Characteristics of Hand Contours Defects , 2016 .

[8]  Chih-Jen Lin,et al.  A comparison of methods for multiclass support vector machines , 2002, IEEE Trans. Neural Networks.

[9]  Patrick Mair,et al.  Goodness-of-Fit Assessment in Multidimensional Scaling and Unfolding , 2016, Multivariate behavioral research.

[10]  N. F. Kirichenko.  Analytical representation of perturbations of pseudoinverse matrices , 1997 .

[11]  Iu. G. Kryvonos,et al.  An Approach to the Determination of Efficient Features and Synthesis of an Optimal Band-Separating Classifier of Dactyl Elements of Sign Language , 2016 .

[12]  Yu. I. Zhuravlev,et al.  A comparison of some approaches to classification problems, and possibilities to construct optimal solutions efficiently , 2014, Pattern Recognition and Image Analysis.

[13]  Iu. V. Krak,et al.  Multidimensional Scaling by Means of Pseudoinverse Operations , 2019, Cybernetics and Systems Analysis.