Abstract
The generative adversarial nets (GANs) are investigated for classification problem. GANs for semi-supervised learning (GAN-SSL) is proposed for complex classification problem. The identification expertise problem is challenging for classification because of complex structure of classes, unbalanced samples and cross-classes. We use semi-supervised learning to solve unbalanced classes problem. Two groups of experiments were carried out. The first group of experiments for the model dataset that consist of classes of points normally distributed about vertices an eight-dimensional hypercube. The second groups of experiments for the petrol identification expertise dataset we get from laboratory of petrol quality. The experiments with good model examples get good quality more than 97%. The analysis of network parameters and generator properties is made. The classification model for petrol identification expertise was created and has 94% quality. In this work we use GAN-SSL classification on petrol identification expertise example, but this classification model can be used for diesel fuel, household chemicals items, different oils and for various other objects.
This is a preview of subscription content, log in via an institution to check access.
Access this chapter
Tax calculation will be finalised at checkout
Purchases are for personal use only
Similar content being viewed by others
References
Goodfellow, I., et al.: Generative adversarial nets. In: 27th Annual Conference on Advances in Neural Information Processing Systems 27, pp. 2672–2680. Curran Associates, Inc., Red Hook (2014)
Salimans, T., Goodfellow, I., Zaremba, W., Cheung, W., Radford, A., Chen, X.: Improved techniques for training GAN. In: Advances in Neural Information Processing Systems, pp. 2234–2242 (2016)
Maksimova, A.: Formal statement of the problem of identification expertise. In: Donetsk Readings 2017: Russian World as a Civilizational Basis for Scientific, pp. 69–70. Educational and Cultural Development of Donbass, Donetsk (2017). (in Russian)
Maksimova, A.: The approach to the construction of information automated systems of identification examination based on machine learning methods. In: International Scientific and Technical Congress Intelligent Systems and Information Technologies, vol. 1, pp. 438–443, Taganrog (2017). (in Russian)
Maksimova, A.: Fuzzy approach to solve pattern recognition problem for automatisation system for identification expertise (example for petrol identification). In. International Scientific Conference Computer Science and Information Technology, Saratov, pp. 256–258 (2016). (in Russian)
Springenberg, J.: Unsupervised and semi-supervised learning with categorical generative adversarial networks (2016). https://arxiv.org/abs/1511.06390
Kingma, D., Rezende, D., Mohamed, S., Welling, M.: Semi-supervised learning with deep generative models. In: Proceedings of the International Conference on Machine Learning, pp. 3581–3589 (2014)
Ding, M., Tang, J., Zhang, J.: Semi-supervised learning on graph with generative adversarial nets. In. Proceedings of the 27th ACM International Conference on Information and Knowledge Management, pp. 913–922 (2018)
Kovalev, V., Kazlouski, S.: Examining the capability of GANs to replace real biomedical images in classification models training (2019). https://arxiv.org/ftp/arxiv/papers/1904/1904.08688.pdf
Ioffe, S., Shegedy, Ch.: Batch normalisation: accelerating deep network training by redusing interval covariance shift. In: ICML 2015, pp. 448–456 (2015)
Saliman, T., Kigma, D.: Weight normalisation: a simple reparametrization to accelerate training of deep neural networks. In: NIPS 2016, pp. 901–909 (2016)
Kigma, D.P., Ba, J.: Adam: a method for stochastic optimization. CoRR abs/1412.6980 (2015). https://arxiv.org/pdf/1412.6980.pdf
Dai, Zh., Yang, Zh., Yang, F., Cohen, W., Salakhutdiov, R.: Good semi-supervised learning that requires a bad GAN. In: 31st Conference on Neural Information Processing Systems (NIPS 2017), Long Beach, CA, USA (2017)
Liu, X., Xiang, X.: How does GAN-based semi-supervised learning work? (2020). https://arxiv.org/abs/1809.00130
Author information
Authors and Affiliations
Corresponding author
Editor information
Editors and Affiliations
Rights and permissions
Copyright information
© 2022 Springer Nature Switzerland AG
About this paper
Cite this paper
Maksimova, A. (2022). Investigation of the GAN-SSL Classifier Properties for Identification Expertise. In: Tuzikov, A.V., Belotserkovsky, A.M., Lukashevich, M.M. (eds) Pattern Recognition and Information Processing. PRIP 2021. Communications in Computer and Information Science, vol 1562. Springer, Cham. https://doi.org/10.1007/978-3-030-98883-8_7
Download citation
DOI: https://doi.org/10.1007/978-3-030-98883-8_7
Published:
Publisher Name: Springer, Cham
Print ISBN: 978-3-030-98882-1
Online ISBN: 978-3-030-98883-8
eBook Packages: Computer ScienceComputer Science (R0)