Abstract
Researchers are currently working on the development of “Explainable Artificial Intelligence” or “Explainable Artificial Intelligence (XAI)”. Such systems are designed to help the user understand the decisions made by the neural network, which will increase confidence in such systems, will allow making more effective decisions based on the results of the system operation. The first results of applying this approach allowed developers and users to study the factors that are used by the neural network to solve a specific problem and what parameters of the neural network need to be changed to improve the accuracy of its work. In addition, studying how neural networks extract, store, and transform knowledge may be useful for the future development of machine learning methods. To overcome this disadvantage of neural networks, it is proposed to consider methods for extracting rules from neural networks, which can become a link between symbolic and connectionistic models of knowledge representation in artificial intelligence. In this paper, we propose a neuro-fuzzy approach to rule extraction using time series forecasting and text recognition as examples.
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
Siau, K., Wang, W.: Building trust in artificial intelligence, machine learning, and robotics. Cut. IT J. 31(2), 47–53 (2018)
Yezhov, A.A., Shumski, S.A.: Neurocomputing and its applications in economics and business. Mephi 224, 10–14 (1998)
Aliev, R.A., et al.: Type-2 fuzzy neural networks with fuzzy clustering and differential evolution optimization. Inf. Sci. 181(9), 1591–1608 (2011). https://doi.org/10.1016/j.ins.2010.12.014
Craven, M., Shavlik, J.: Extracting tree-structured representations of trained networks. Adv. Neural Inf. Process. Syst. 8, 24–30 (1995)
Gridin, V.N., Solodovnikov, V.I., Evdokimov, I.A., Filippkov, S.V.: Building decision trees and extracting rules from trained neural networks. Artif. Intell. Decis. Mak. 4, 26–33 (2013)
Shevelev, O.G., Petrakov, A.V.: Classification of texts using decision trees and feedforward neural networks. Bull. Tomsk State Univ. 290 (2006)
Yarushev, S.A., Averkin, A.N.: Time series analysis based on modular architectures of neural networks. Procedia Comput. Sci. 123, 562–567 (2018). https://doi.org/10.1016/j.procs.2018.01.085
Acknowledgements
This research was performed in the framework of the state task in the field of scientific activity of the Ministry of Science and Higher Education of the Russian Federation, project “Development of the methodology and a software platform for the construction of digital twins, intellectual analysis and forecast of complex economic systems”, grant no. FSSW-2020-0008.
Author information
Authors and Affiliations
Editor information
Editors and Affiliations
Rights and permissions
Copyright information
© 2022 The Author(s), under exclusive license to Springer Nature Switzerland AG
About this paper
Cite this paper
Averkin, A., Yarushev, S. (2022). Explainable Artificial Intelligence: Rules Extraction from Neural Networks. In: Aliev, R.A., Kacprzyk, J., Pedrycz, W., Jamshidi, M., Babanli, M., Sadikoglu, F.M. (eds) 11th International Conference on Theory and Application of Soft Computing, Computing with Words and Perceptions and Artificial Intelligence - ICSCCW-2021. ICSCCW 2021. Lecture Notes in Networks and Systems, vol 362. Springer, Cham. https://doi.org/10.1007/978-3-030-92127-9_17
Download citation
DOI: https://doi.org/10.1007/978-3-030-92127-9_17
Published:
Publisher Name: Springer, Cham
Print ISBN: 978-3-030-92126-2
Online ISBN: 978-3-030-92127-9
eBook Packages: Intelligent Technologies and RoboticsIntelligent Technologies and Robotics (R0)