Abstract
The method of code description of classes, which is a development of the ECOC (error-correcting output codes) method, is grounded theoretically. The main difference is as follows: a multiset of code descriptions of its training objects is used instead of one object code. It is shown that under certain conditions the two methods are equivalent. Ways for improving the recognition quality if code descriptions are used are shown.
Similar content being viewed by others
References
Yu. I. Zhuravlev, “Correct algebras over sets of incorrect (heuristic) algorithms. I,” Cybern. 13 (4), 489–497 (1977).
Yu. I. Zhuravlev, “Correct algebras over sets of incorrect (heuristic) algorithms. II,” Cybern. 13 (6), 814–821 (1977).
C. Cortes and V. Vapnik, “Support–vector networks,” Mach. Learn. 20 (3), 273–297 (1995).
V. A. Kuznetsov, O. V. Sen’ko, A. V. Kuznetsova, L. P. Semenova. A. V. Aleshchenko, T. B. Gladysheva, and A. V. Ivshina, “Recognition of fuzzy systems by the method of statistically weighted syndromes and its application for the immunohematological characterization of norm and chronic pathology,” Khim. Fizika (Chem. Phys.) 15 (1), 81–100 (1996) [in Russian].
T. G. Dietterich and G. Bakiri, “Solving multiclass learning problems via error–correcting output codes,” J. Artif. Intell. Res. 2, 263–286 (1995).
E. Allwein, R. Shapire, and Y. Singer, “Reducing multi–class to binary: A unifying approach for margin classifiers,” J. Mach. Learn. Res. 1 (1), 113–141 (2000).
A. A. Dokukin, V. V. Ryazanov, and O. V. Shut, “Multilevel models for solution of multiclass recognition problem,” Pattern Recogn. Image Anal.: Adv. Math. Theory Appl. 26 (3), 461–473 (2016).
A. A. Dokukin, V. V. Ryazanov, and O. V. Shut, “Multilevel models for pattern recognition tasks with multiple classes,” Inform. Primen. (Inf. Appl.) 11 (1), 69–78 (2017).
M. Lichman, UCI Machine Learning Repository (University of California, School of Information and Computer Science, Irvine, CA, 2013). http://archive.ics.uci.edu/ml
K. D. Schmidt, “On the covariance of monotone functions of a random variable,” Manuscript in Dresdner Schriften zur Versicherungsmathematik (Technical University of Dresden, 2003), pp. 1–3. Available at https://www.math.tu–dresden. de/sto/schmidt/dsvm/dsvm2003–4.pdf
Author information
Authors and Affiliations
Corresponding author
Additional information
Aleksandr Aleksandrovich Dokukin. Born in 1980, senior scientist of Federal Research Center “Computer Science and control” of the Russian Academy of Sciences. Graduated with honors from Faculty of Computational Mathematics and Cybernetics of Lomonosov Moscow State University in 2002. In 2005, received post-graduate degree from the same faculty. Since 2008, candidate of physical and mathematical sciences. Since 2000 has worked at Dorodnicyn Computing Centre of the Russian Academy of Sciences (later known as part of FRC CSC RAS). Fields of interests: pattern recognition, data analysis. Author or co-author of 81 scientific works.
Rights and permissions
About this article
Cite this article
Dokukin, A.A. Method of Code Description of Classes for Solving Multi-Class Problem. Pattern Recognit. Image Anal. 28, 688–694 (2018). https://doi.org/10.1134/S1054661818040077
Received:
Published:
Issue Date:
DOI: https://doi.org/10.1134/S1054661818040077