High Generalization Capability Artificial Neural Network Architecture Based on RBF-Network
This paper describes the issue of error level fluctuations due to training set shrinking in RBF-networks. An architecture of artificial neural network (ANN) based on RBF-network is presented with a learning algorithm to train it. The presented architecture is multi-layer, unlike original RBF-network thus has a potential in deep learning. Numeric results lead to a conclusion about error level fluctuations being significantly lower for the presented architecture compared to RBF-network in case of training set shrinking. This displays a greater generalization ability of the presented architecture. The paper contains an application of ANN to the task of restoring the dielectric parameters for subject placed in waveguide.
KeywordsArtificial neural network Neural network learning algorithm RBF neural network
- 1.Sermanet, P., Eigen, D., Zhang, X., Mathieu, M., Fergus, R., LeCun, Y.: OverFeat: integrated recognition, localization and detection using convolutional networks. In: International Conference on Learning Representations (ICLR2014), CBLS, (arXiv:1312.6229) (2014)
- 7.Abrosimov, M.A., Brovko, A.V.: Criteria for minimal RBF artificial neural network’s center vectors set definition. Neirokomputery: razrabotka, primenenie. (3), 50–54 (2018). (in Russian)Google Scholar
- 9.Craddock, R.J., Warwick, K.: Multi-layer radial basis function networks. an extension to the radial basis function. In: IEEE Proceedings of International Conference on Neural Networks (ICNN 1996), vol. 2, pp. 700–705 (1996). https://doi.org/10.1109/ICNN.1996.548981
- 10.Chao, J., Hoshino, M., Kitamura, T., Masuda, T.: A multilayer RBF network and its supervised learning. In: Proceedings of IEEE International Joint Conference on Neural Networks, IJCNN 2001 (Cat. No. 01CH37222), vol. 3, pp. 1995–2000 (2001). https://doi.org/10.1109/IJCNN.2001.938470