Научная электронная библиотека ULRICH'S Periodicals Directory

Сибирский государственный университет
науки и технологий имени академика М.Ф. Решетнева

Сибирский аэрокосмический журнал
ISSN 2712-8970 (Print) ISSN 2782-5760 (on-line)

Сведения об образовательной организации

UDK 519.87
SELF-CONFIGURING HYBRID EVOLUTIONARY ALGORITHM FOR FUZZY CLASSIFIER DESIGN WITH ACTIVE LEARNING FOR UNBALANCED DATASETS
V. V. Stanovov, O. E. Semenkina
Siberian State Aerospace University named after academician M. F. Reshetnev 31, Krasnoyarsky Rabochy Av., Krasnoyarsk, 660014, Russian Federation E-mail: vladimirstanovov@yandex.ru
The paper describes an active training example selection for a self-configured hybrid evolutionary algorithm for fuzzy rule bases design for classification problems. This method is related to instance selection methods, which allow not only decreasing of required computational recourses, but also increasing the quality of the obtained classifiers. The method changes the probabilities of instances which are selected into the training subsample depending on how good they are classified by the algorithm. After several generations the sample is changed and probabilities are recalculated. Those instances which were not used before and those which were misclassified by the algorithm had higher probabilities of getting into the training sample. The probabilities of instance selection were calculated using a procedure similar to proportional selection in the genetic algorithm. The idea of training instance selection described here was implemented for the fuzzy classifiers forming. This algorithm uses the combination of Pittsburg and Michigan approach for fuzzy rule base design with fixed terms, and the Michigan approach is used together with the mutation operator. The size of the rule base is not fixed, and may change during the algorithm run, and a corresponding class number and the rule weight were calculated heuristically for every rule. Moreover, the algorithm uses an initialization procedure that uses instances from the sample to generate more accurate rules. In the Michigan part the operators of adding rules, deleting rules and replacing rules has been implemented. The creation of new rules could be performed by genetic approach – using the existing rules, and heuristically – using those instances which were misclassified. The efficiency of the algorithm was shown on a set of complex classification problems with several classes, as an efficiency measure the overall accuracy and the average accuracy among classes was used.
fuzzy classification system, active learning, evolutionary algorithm, self-configuration, unbalanced data
References

  1. Spinks N., Silburn N. L. J., Birchall D. W. Making it all work: The engineering graduate of the future. A UK perspective, European Journal of Engineering Education, 2008, no. 32(3), p. 325–335.

  2. Pol A. P., Moreno J. J. M., Oliver M. P. Generic skills in higher education. Comparative study of the views of employers and academics. Psicothema, 2009,
    no. 21(3), p. 433–438.

  3. Badcock P. B. T., Pattison P. E., Harris K. L. Developing generic skills through university study: A study of arts, science and engineering In Australia, Higher Education, 2010, 60(4), p. 441–458.

  4. Croft A., Ward J. A. Modern and interactive approach to learning engineering mathematics. British Journal of Educational Technology, 2001, no. 32(2),
    p. 195–207.

  5. Shershneva V. A. Formirovanie matematicheskoy kompetentnosty studentov inzhenernogo vuza na osnove poliparadigmal’nogo podkhoda [Formation of mathematical competence of students of engineering university based on polyparadigmatic approach]. Krasnoyarsk, SibSAU Publ., 2011, 268 p.

  6. Shershneva V. A. Kak otsenit’ mezhdistsiplinarnye kompetentsii studentov [How to evaluate the interdisciplinary competence of students]. Vyssheye obrazovanie v Rossii. 2007, no. 10, p. 48–50 (In Russ.).

  7. Bashmakov A. I., Bashmakov I. A. Razrabotka komputernyh uchebnikov I obuchayushyh system [Design of computer textbooks and training systems] Moscow, Filin Publ., 2003, 616 p.

  8. Zykova T. V., Kytmanov A. A., Tsibulsky G. M., Shershneva V. A. [Learning mathematics in the environment Moodle for example of e-learning course]. Vestnik KGPU. 2012, no. 1, p. 60–63 (In Russ.).

  9. Gafurova N. V., Osipova S. I. [On the implementation of psycho-pedagogical learning objectives in the information educational environment]. Siberian pedagogical journal. 2010, no. 1, p. 117–124 (In Russ.).

  10. Belyaev M. I. [From the experience of creating electronic textbooks]. Vestnik RUDN, seryia Informatizatsiya obrazovania.2009, no. 1, p. 11–20 (In Russ.).

  11. Zykova T. V., Kytmanov A. A., Sidorova G. M., Shershneva V. A. [About teaching materials for e-learning course of mathematical analysis, developed
    on the basis of the polyparadigmatic approach]. Vestnik KGPU. 2012, no. 4, p. 109–113 (In Russ.).

  12. Avetisyan D. D. Аветисян Д. Д. Obrazovatel’ny content dlya distantsionnogo obucheniya [Educational content for distance learning]. Prepodavatel'. XXI vek. 2009, no. 1, p. 51–59 (In Russ.).

  13. Mineev N. S. [Electronic textbook is a modern means of teaching students]. Yaroslavsky pedagogichesky vestnik. 2012, vol. 2, no. 2, p. 221–224 (In Russ.).

  14. Zykova T. V., Sidorova G. M., Shershneva V. A., Tsibulsky G. M. [An experience of the use of web-based environment Moodle for learning of mathematics students of engineering university based polyparadigmatic
    approach]. Informatika i obrazovanie. 2013, no. 5 (244), p. 37–40 (In Russ.).

  15. Zykova T. V., Karnaukhova O. A., Sidorova G. M., Shershneva V. A. [Features of e-learning mathematics of engineering university students]. Vestnik KGPU. 2014, no. 3 (29), p. 55–61 (In Russ.).


Stanovov Vladimir Vadimovich – postgraduate student, Siberian State Aerospace University named after academician M. F. Reshetnev. Е-mail: vladimirstanovov@yandex.ru

Semenkina Olga Ernestovna – Dr. Sc., Professor, Professor of the Higher Mathematics Department, Siberian State Aerospace University named after academician M. F. Reshetnev. Е-mail: semenkina.olga@mail.ru