Abstract
We propose a deep learning algorithm that breaks with the paradigm of weights and activation functions, CEFYDRA, a network of cluster-first fuzzy-based regression algorithms. In this paper, we cover the generalization of CEFYDRA to multilayered deep architectures. First, we provide the three laws that make this generalization possible: Displacement, Substitution and Composition. Then, we obtain the update formulas for the parameters of deep hidden layers. Finally, we show the pseudocode for prediction and update. We also briefly mention the reasons to believe that this algorithm, named CEFYDRA, is explainable and has the ability for plastic reorganization.
Access this chapter
Tax calculation will be finalised at checkout
Purchases are for personal use only
References
Yang, Z., Zhang, A., Sudjianto, A.: GAMI-Net: an explainable neural network based on generalized additive models with structured interactions. Pattern Recognit. 120, 108192 (2021)
Lee, Y.: Extraction of competitive factors in a competitor analysis using an explainable neural network. Neural Process. Lett. 53(3), 1979ā1994 (2021). https://doi.org/10.1007/s11063-021-10499-6
Fauvel, K., Lin, T., Masson, V., Fromont, Ć., Termier, A.: XCM: an explainable convolutional neural network for multivariate time series classification. Mathematics 9(23), 3137 (2021)
Sasaki, H., Hidaka, Y., Igarashi, H.: Explainable deep neural network for design of electric motors. IEEE Trans. Magn. 57(6), 1ā4 (2021)
Valizadegan, H. et al.: ExoMiner: A Highly Accurate and Explainable Deep Learning Classifier that Validates 301 New Exoplanets. arXiv:2111.10009 (2021)
Kim, M.S., Yun, J.P., Park, P.: An explainable convolutional neural network for fault diagnosis in linear motion guide. In IEEE Trans. Ind. Inform. 17(6), 4036ā4045 (2021)
Gallego, V., RĆos Insua, D.: Current advances in neural networks. Annu. Rev. Stat. Appl. 9(1), 197ā222 (2022)
Samek, W., Montavon, G., Lapuschkin, S., Anders, C.J., MĆ¼ller, K.-R.: Explaining deep neural networks and beyond: a review of methods and applications. Proc. IEEE 109(3), 247ā278 (2021)
Viana, J., Cohen, K., Ralescu, A., Ralescu, S., Kreinovich, V.: Localized learning: a possible alternative to current deep learning techniques. In: Castillo, O., Melin, P. (eds.) New Perspectives on Hybrid Intelligent System Design Based on Fuzzy Logic, Neural Networks and Metaheuristics, Springer (2023)
ViaƱa, J., Cohen, K.: Fuzzy-based, noise-resilient, explainable algorithm for regression. In: Rayz, J., Raskin, V., Dick, S., Kreinovich, V. (eds.) NAFIPS 2021. LNNS, vol. 258, pp. 461ā472. Springer, Cham (2022). https://doi.org/10.1007/978-3-030-82099-2_42
ViaƱa, J., Ralescu, S., Cohen, K., Ralescu, A., Kreinovich, V.: Extension to multi-dimensional problems of a fuzzy-based explainable and noise-resilient algorithm. In: Proceedings of the 14th International Workshop on Constraint Programming and Decision Making CoProdā2021, Szeged, Hungary (2021)
ViaƱa, J., Ralescu, S., Cohen, K., Ralescu, A., Kreinovich, V.: Why cauchy membership functions: reliability. Adv. Artif. Intell. Mach. Learn. 2(2), 385ā393 (2022)
ViaƱa, J., Ralescu, S., Cohen, K., Ralescu, A., Kreinovich, V.: Why cauchy membership functions: efficiency. Adv. Artif. Intell. Mach. Learn. 1(1), 81ā88 (2021)
Acknowledgements
The project that generated these results was supported by a grant from the āla Caixaā Banking Foundation (ID 100010434), whose code is LCF/BQ/AA19/11720045.
Author information
Authors and Affiliations
Corresponding author
Editor information
Editors and Affiliations
Rights and permissions
Copyright information
Ā© 2023 The Author(s), under exclusive license to Springer Nature Switzerland AG
About this paper
Cite this paper
ViaƱa, J., Ralescu, S., Kreinovich, V., Ralescu, A., Cohen, K. (2023). Multiple Hidden Layered CEFYDRA: Cluster-First Explainable Fuzzy-Based Deep Self-reorganizing Algorithm. In: Dick, S., Kreinovich, V., Lingras, P. (eds) Applications of Fuzzy Techniques. NAFIPS 2022. Lecture Notes in Networks and Systems, vol 500. Springer, Cham. https://doi.org/10.1007/978-3-031-16038-7_30
Download citation
DOI: https://doi.org/10.1007/978-3-031-16038-7_30
Published:
Publisher Name: Springer, Cham
Print ISBN: 978-3-031-16037-0
Online ISBN: 978-3-031-16038-7
eBook Packages: Intelligent Technologies and RoboticsIntelligent Technologies and Robotics (R0)