Radial Basis Function (RBF) neural networks are multi-layer feed forward networks widely used for pattern recognition, function approximation and data interpolation problems in areas such as time series analysis, image processing and medical diagnosis. Rather than using the sigmoid activation function as in back propagation networks, the hidden neurons in a RBF neural network use a Gaussian or some other radial basis function. As part of a theoretical and experimental study of RBF neural networks, in this paper we adopt a Fock spaces-based approach, and show how it can be applied to the modeling a RBF neural network. Specifically, we are working to explore and to understand the new theoretical and operational aspects of this kind of neural network from a quantum perspective.
ISSN: 1792-6939 (Online)