1. Urenda, J.C., Csiszar, O., Csiszar, G., Dombi, J., Kosheleva, O., Kreinovich, V., Eigner, G.: Why squashing functions in multi-layer neural networks. In: Proceedings of the 2020 IEEE International Conference on Systems, Man, and Cybernetics SMC’2020, Toronto, Canada, pp. 296–300 (2020). Accessed from 11–14 Oct 2020
2. Goodfellow, I., Bengio, Y., Courville, A.: Deep Learning. MIT Press, Cambridge, Massachusetts (2016)
3. Fuentes, O., Parra, J., Anthony, E., Kreinovich, V.: Why rectified linear neurons are efficient: a possible theoretical explanations. In: Kosheleva, O., Shary, S., Xiang, G., Zapatrin, R. (eds.) Beyond Traditional Probabilistic Data Processing Techniques: Interval, Fuzzy, etc, pp. 603–613. Methods and Their Applications, Springer, Cham, Switzerland (2020)
4. Kreinovich, V., Kosheleva, O.: Deep learning (partly) demystified. In: Proceedings of the 2020 4th International Conference on Intelligent Systems, Metaheuristics & Swarm Intelligence ISMSI’2020, Thimpu, Bhutan (2020). Accessed from 18–19 Apr 2020
5. Csiszár, O., Csiszár, G., Dombi, J.: Interpretable Neural Networks Based on Continuous-Valued Logic and Multicriterion Decision Operators (2020). arXiv:1910.02486v2, posted on February 7, 2020