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 004.94
OPTIMISATION OF CONVOLUTIONAL NEURAL NETWORK STRUCTURE WITH SELF-CONFIGURING EVOLUTIONARY ALGORITHM IN ONE IDENTIFICATION PROBLEM
D. V. Fedotov*, E. A. Popov, V. A. Okhorzin
Reshetnev Siberian State Aerospace University 31, Krasnoyarsky Rabochy Av., Krasnoyarsk, 660037, Russian Federation *E-mail: fedotov.dm.v@gmail.com
Computer technology development has opened great opportunities for researchers in the field of dynamic systems optimization offering new algorithms and ways of their combination into complex and powerful intelligent information technologies. A great interest in the field of machine learning has been given to dynamic objects identification and pat-tern recognition in computer vision systems in particular recently. Computer vision problems arise in various fields: industry, security and surveillance systems, data acquisition and processing systems, computer-human interaction sys-tems, etc. Neural networks are widely and successfully used for solving machine learning tasks. Neural networks are computer models based on network of nerve cells of a living organism. Using classical neural network causes major difficulties for solving computer vision tasks as they require significant computational and/or time resources for learn-ing as well as they lose important information about topology of the original data. For this kind of tasks a special type of neural network called convolutional neural network was developed. Convolutional neural network (CNN) is the part of subfield of machine learning called deep learning. CNN is used as the main technology in this paper. It allows to build complex hierarchies of features and perform objects identification based on them. Using of the pooling layers provides invariance to size of the image and concept of parameters sharing can significantly reduce the number of pa-rameters that have to be adjusted and therefore save computational costs and time. Standard training method for neural network (back propagation) has certain weaknesses that can be partially eliminated by using a evolutionary optimiza-tion algorithm. The quality of the solution depends on the neural network structure, which can also be adjusted using evolutionary algorithms. In this paper, the self-configuring genetic algorithm SelfCGA is used for CNN’s structure and weighting coefficients adjustment. Proposed system is tested on the task of age identification based on the person’s photo.
Keywords: convolutional neural network, evolutionary algorithms, self-configuring, computer vision, machine learning, identification.
References

References

 

  1. LeCun Y., Bottou L., Bengio Y., Haffner P. Gradient-Based Learning Applied to Document Recognition. Proceedings of the IEEE, 1998, No. 86(11), P. 2278–2324.
  2. Bottou L., LeCun Y., Bengio Y. Global training of document processing systems using graph transformer networks. Proceedings of Computer Vision and Pattern Recognition, Puerto-Rico, 1997, P. 490–494.
  3. LeCun Y., Bottou L., Bengio Y. Reading checks with graph transformer networks. International Conference on Acoustics, Speech, and Signal Processing, 1997, Vol. 1, P. 151–154.
  4. LeCun Y., Jackel L. D., Bottou L et al. Comparison of learning algorithms for handwritten digit recognition. International Conference on Artificial Neural Networks, 1995, P. 53–60.
  5. LeCun Y., Bengio Y. Convolutional Networks for Images, Speech, and Time-Series. The Handbook of Brain Theory and Neural Networks, MIT Press, 1995, P. 255–258.
  6. Bengio Y., Goodfellow I. J., Courville A. Deep Learning Book in preparation. MIT Press, 2015, P. 173–190.
  7. Wright A. Genetic algorithms for real parameter optimization. Foundations of Genetic Algorithms, 1991, P. 205–218.
  8. Eiben A. E., Smith J. E. Introduction to evolutionary computing. Springer, Berlin, Germany 2003, 300 p.
  9. Sergienko A. B. Geneticheskiy algoritm. Standart. Versiya 3.12 [Genetic algorithm. Standard. Version 3.12], Krasnoyarsk, SibGAU Semenkin E., Semenkina M. Self-configuring genetic algorithm with modified uniform crossover operator. Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), 2012, 7331 LNCS (Part 1), P. 414–421.
  10. Semenkin E., Semenkina M. Self-configuring genetic algorithm with modified uniform crossover operator. Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics). 2012. 7331 LNCS (PART 1). P. 414–421.
  11. Gomez J. Self Adaptation of Operator Rates Evolutionary Algorithms. Genetic and Evolutionary Computation, GECCO 2004, Lecture Notes in Computer Science, 2004, Vol. 3102, P. 1162–1173.
  12. Semenkina M., Semenkin E. Hybrid Self-configuring Evolutionary Algorithm for Automated
    Design of Fuzzy Classifier. Advances in Swarm Intelligence ICSI, 2014, Part 1, LNCS 8794, P. 310–317.
  13. Burakov S. V., Semenkin E. S. Solving variational and Cauchy problems with self-configuring genetic programming algorithms. International Journal of Innovative Computing and Applications, 2013, Vol. 5, Iss. 3, P. 152–162.
  14. Semenkin E., Semenkina M. Classifier Ensembles Integration with Self-Configuring Genetic Programming Algorithm. Adaptive and Natural Computing Algorithms. Lecture Notes in Computer Science, 2013, Vol. 7824, P. 60–69.
  15. Fedotov D. V., Semenkin E. S. [On forecasting economic indexes by means of neuroevolutionary models]. Vestnik SibGAU, 2014, No. 5 (57), P. 299–304 (In Russ.).
  16. Fedotov D. V. On financial time series prediction using neural network models and evolutionary algorithms. Materials of the Third International Workshop on Mathematical Models and its Applications (IWMMA-2014), 2014, P. 1–3.
  17. Fedotov D. V. [On using convolutional neural networks for pattern recognition tasks]. Prospekt Svobodnyy-2015 : materialy nauch. konf. [Svobodnyy avenue 2015: Materials of scientific conference], 2015, P. 56–57 (In Russ.).
  18. Werbos P. J. Beyond regression: New tools for prediction and analysis in the behavioral sciences. Harvard University, Cambridge, MA, 1974, 906 p.

Fedotov Dmitrii Valer`evich – engineer of RD department, Reshetnev Siberian State Aerospace University. E-mail: fedotov.dm.v@gmail.com.

Popov Eugene Aleksandrovich – Dr. Sc., professor, professor of System analysis and operation research department, Reshetnev Siberian State Aerospace University. E-mail: epopov@bmail.ru.

Okhorzin Vladimir Afanas’evich – Dr. Sc., professor, professor of Applied mathematics department, Reshetnev Siberian State Aerospace University. Е-mail: okhorzin@mail.ru.