Title :
Generalized CNN: Potentials of a CNN with non-uniform weights
Author_Institution :
Dipartimento di Ingegneria Elettronica, Roma Univ., Italy
Abstract :
A generalization of the cellular neural network (CNN) paradigm is obtained by removing the uniformly constraint on weight values. Such generalized CNNs are capable of new tasks, such as function approximation or associative memory. A stability analysis of these networks is presented. Adaptation and application of a gradient descent learning algorithm is then discussed
Keywords :
constraint handling; function approximation; generalisation (artificial intelligence); learning (artificial intelligence); neural nets; stability; CNN; associative memory; cellular neural network; function approximation; generalization; gradient descent learning; non-uniform weights; stability analysis; uniformly constraint; weight values; Asymptotic stability; CADCAM; Cellular neural networks; Cloning; Computer aided manufacturing; Convolution; Function approximation; Image processing; Integrated circuit layout; Neural networks;
Conference_Titel :
Cellular Neural Networks and their Applications, 1992. CNNA-92 Proceedings., Second International Workshop on
Conference_Location :
Munich
Print_ISBN :
0-7803-0875-1
DOI :
10.1109/CNNA.1992.274342
