Title :
A convergence theorem for incremental learning with real-valued inputs
Author :
Gordon, Mirta B.
Author_Institution :
CEA, Centre d´´Etudes Nucleaires, de Grenoble, France
Abstract :
We present a convergence theorem for incremental learning algorithms, valid for real-valued input patterns. The upper bound to the number of hidden units is equal to P-1, where P is the number of patterns in the training set
Keywords :
convergence of numerical methods; learning (artificial intelligence); neural nets; pattern classification; set theory; convergence theorem; hidden units; incremental learning; neural networks; parity machine; pattern classification; real-valued inputs; upper bound; Convergence; Machine learning; Neural networks; Neurons; Radiofrequency interference; Upper bound;
Conference_Titel :
Neural Networks, 1996., IEEE International Conference on
Conference_Location :
Washington, DC
Print_ISBN :
0-7803-3210-5
DOI :
10.1109/ICNN.1996.548922
