DocumentCode :
2624346
Title :
Asymptotic behaviors of simulated annealing and mean-field approximate annealing
Author :
Matsuba, Ikuo ; Masui, Hironari
Author_Institution :
Hitachi Ltd., Kawasaki, Japan
fYear :
1991
fDate :
18-21 Nov 1991
Firstpage :
923
Abstract :
The asymptotic behaviors of simulated annealing are investigated both theoretically and numerically for the fully connected Hopfield neural network. The energy using ordinary simulated annealing is found to scale as 1/log(t), while the energy using the mean-field approximate annealing method proposed in the present work shows a faster scaling property described by 1/t1/2, where t is an iterative time step
Keywords :
iterative methods; neural nets; simulated annealing; asymptotic behaviors; fully connected Hopfield neural network; iterative time step; mean-field approximate annealing; scaling property; simulated annealing; Boltzmann distribution; Cooling; Costs; Hopfield neural networks; Iterative methods; Laboratories; Neurons; Optimization methods; Simulated annealing; Temperature;
fLanguage :
English
Publisher :
ieee
Conference_Titel :
Neural Networks, 1991. 1991 IEEE International Joint Conference on
Print_ISBN :
0-7803-0227-3
Type :
conf
DOI :
10.1109/IJCNN.1991.170518
Filename :
170518
Link To Document :
https://search.ricest.ac.ir/dl/search/defaultta.aspx?DTC=49&DC=2624346
بازگشت