Universal approximation of fully complex feed-forward neural networks

ICASSP, IEEE International Conference on Acoustics, Speech and Signal Processing - Proceedings

Volumn 1, Issue , 2002, Pages

  Kim, Taehwan ^a   Adali, Tülay ^a  

^a University of Marylandw Baltimore County   (United States)

Author keywords

[No Author keywords available]

Indexed keywords

APPROXIMATION THEORY; CONFORMAL MAPPING; CONVERGENCE OF NUMERICAL METHODS; FUNCTIONS; PROBABILITY; SET THEORY;

ELEMENTARY TRANSCENDENTAL FUNCTIONS; NONLINEAR ACTIVATION FUNCTION;

FEEDFORWARD NEURAL NETWORKS;

EID: 0036289981     PISSN: 15206149     EISSN: None     Source Type: Conference Proceeding    
DOI: 10.1109/icassp.2002.5743956     Document Type: Conference Paper

References (16)

1. Generalization of neural network to the complex plane
2. The complex backpropagation algorithm
3. Non linear satellite radio links equalized using blind neural networks
4. Complex backpropagation
5. Continuous complex-valued back-propagation learning
6. Complex-valued neural networks with adaptive splines activation function
7. The use of Möbius transformations in neural networks and signal processing
8. Fully complex backpropagation for constant envelope signal processing
9. Complex backpropagation neural network using elementary transcendental activation functions
10. Approximation by superpositions of a sigmoidal function
11. Multilayer feedforward networks are universal approximators