Informatics and Applications
2017, Volume 11, Issue 1, pp 69-78
MULTILEVEL MODELS FOR PATTERN RECOGNITION TASKS WITH MULTIPLE CLASSES
- A. A. Dokukin
- V. V. Ryazanov
- O. V. Shut
Abstract
The problem of choosing binary subtasks for recognition tasks with multiple classes is considered from the points of view of the algebraic and logical approaches to recognition. The limits of their applicability were studied theoretically. The sufficient condition of correctness of algorithms is stated as a result of research of dependency between the first and the second level algorithms. Additionally, the paper proves that the object resolution method is applicable to constructing new objects using the precedent information. As an applied result, two modifications of the ECOC (error-correcting output codes) method are proposed. The first one is based on optimization of the binary subtasks set. The second one develops ideas of the fuzzy object resolution method with classes described by multisets of codes of their precedents. The proposed modifications make it possible to increase the initial method's quality in various situations, which is demonstrated by the example of model and real-world tasks.
[+] References (17)
- Zhuravlev, Yu. I. 1977. Korrektnye algebry nad mnozhestvom nekorrektnykh (evristicheskikh) algoritmov I [Correct algebras over sets of incorrect (heuristic) algorithms I]. Kibernetika [Cybernetics] 4:14-21.
- Zhuravlev, Yu. I. 1977. Korrektnye algebry nad mnozhestvom nekorrektnykh (evristicheskikh) algoritmov II [Correct algebras over sets of incorrect (heuristic) algorithms II]. Kibernetika [Cybernetics] 6:21-27.
- Cortes, C., and V. Vapnik. 1995. Support-vector networks. Mach. Learn. 3(20):273-297.
- Kuznetsov, V. A., O. V. Senko, A. V. Kuznetsova, L. P Se-menova, A. V. Aleshchenko, T. B. Gladysheva, and
A. V. Ivshina. 1996. Raspoznavanie nechetkikh sistem po metodu statisticheski vzveshennykh sindromov i ego primenenie dlya immunogematologicheskoy kharakteri-
stiki normy i khronicheskoy patologii [Recognition of fuzzy systems by method of statistically weighed syn-dromes and its using for immunological and hematological norm and chronic pathology]. Khim. Fiz. 15(1):81-100.
- Knerr, S., L. Personnaz, and G. Dreyfus. 1990. Singlelayer learning revisited: A stepwise procedure for building and training neural network. Neurocomputing: Algorithms, architectures and applications. Eds. F F. Soulie and J. Her- ault. NATO ASI subser. F. Berlin-Heidelberg: Springer- Verlag. 68:41-50.
- Dietterich, T. G., and G. Bakiri. 1995. Solving multiclass learning problems via error-correcting output codes. J. Artif. Intell. Res. 2:263-286.
- Allwein, E., R. Shapire, and Y. Singer. 2000. Reducing multi-class to binary: A unifying approach for margin classifiers. J. Mach. Learn. Res. 1(1):113-141.
- D'yakonov, A. G. 2005. Algebra over estimation algorithms: The minimal degree of correct algorithms. Comp. Math. Math. Phys. 45(6):1095-1106.
- Dokukin, A. A. 2001. The construction of a recognition algorithm in the algebraic closure. Comp. Math. Math. Phys. 41(12):1907-1911.
- Dokukin, A., V. Ryazanov, and O. Shut. 2016. Multilevel models for solution of multiclass recognition problems. Pattern Recognition Image Anal. 26(3):461-473.
- Giarratano, J. C., and G. D. Riley. 2004. Expert systems: Principles and programming. Boston, MA: PWS Publ. Co. 856 p.
- Krasnoproshin, V.V., and V.A. Obraztsov. 1998. Raspo- znavanie s obucheniem kak zadacha vybora [Supervised recognition as selection problem]. Tsifrovaya obrabotka izobrazheniy [Digital image processing]. Minsk: ITK. 8094.
- Ablamejko, S.V., V. V. Krasnoproshin, and V.A. Obraz-tsov. 2011. Modeli i tehnologii raspoznavaniya obra-
zov s prilozheniem v intellektual'nom analize dannykh [Models and technologies of pattern recognition with ap-plication to data mining]. Vestnik BGU. Ser. 1, Fizika, Matematika, Informatika [BSU Herald. Ser. 1, Physics, Mathematics, Informatics] 3:62-72.
- Kaufman, A. 1975. Introduction to the theory offuzzy sub-sets. New York, NY: Academic Press. 432 p.
- Lee, R. C. T. 1972. Fuzzy logic and the resolution principle. J.ACM 19(1):109-119.
- Lichman, M. 2013. UCI machine learning repository. Irvine, CA: University of California, School of Information and Computer Science. Available at: http://archive.ics.uci.edu/ml (accessed March 3, 2017).
- Pedregosa, F., G. Varoquaux, A. Gramfort, V. Michel, B. Thirion, O. Grisel, M. Blondel, P. Prettenhofer, R. Weiss, V. Dubourg, J. Vanderplas, A. Passos, D. Cournapeau, M. Brucher, M. Perrot, and E. Duchesnay. 2011. Scikit-learn: Machine learning in Python. J. Mach. Learn. Res. 12:2825-2830.
[+] About this article
Title
MULTILEVEL MODELS FOR PATTERN RECOGNITION TASKS WITH MULTIPLE CLASSES
Journal
Informatics and Applications
2017, Volume 11, Issue 1, pp 69-78
Cover Date
2017-02-30
DOI
10.14357/19922264170106
Print ISSN
1992-2264
Publisher
Institute of Informatics Problems, Russian Academy of Sciences
Additional Links
Key words
classification; multiclass task; ECOC; multilevel method; correctness; algebraic approach; logical approach; code class description
Authors
A. A. Dokukin  , V. V. Ryazanov  ,
and O. V. Shut 
Author Affiliations
Federal Research Center “Computer Sciences and Control” of the Russian
Academy of Sciences, 44-2 Vavilov Str.,Moscow 119333, Russian Federation
Moscow Institute of Physics and Technology, 9 Institutskiy Per., Dolgoprudny, Moscow Region 141700, Russian Federation
Belarusian State University, 4 Nezavisimosti Av., Minsk 220030, Republic of Belarus
|