Title :
Efficient parallel implementation of Kolmogorov superpositions
Author_Institution :
Inst. of Comput. Sci., Acad. of Sci. of the Czech Republic, Prague, Czech Republic
fDate :
6/24/1905 12:00:00 AM
Abstract :
We analyze serial and parallel implementation of the learning algorithm based on Kolmogorov superposition theorem. Theoretical time complexity estimates are compared and parallel speedup is determined. Practical experiments show that the speedup in the order of 2n, where n is the input dimension, is achievable for real parallel environments (such as clusters of workstations)
Keywords :
computational complexity; learning (artificial intelligence); parallel algorithms; Kolmogorov superpositions; clusters of workstations; learning algorithm; parallel environments; parallel implementation; parallel speedup; time complexity; Algorithm design and analysis; Clustering algorithms; Computer networks; Computer science; Estimation theory; Neural networks; Quantum computing; Workstations; Zinc;
Conference_Titel :
Neural Networks, 2002. IJCNN '02. Proceedings of the 2002 International Joint Conference on
Conference_Location :
Honolulu, HI
Print_ISBN :
0-7803-7278-6
DOI :
10.1109/IJCNN.2002.1007669
