Seeking Global Minima

✍ Scribed by Wan Ahmad Tajuddin Wan Abdullah

• Book ID: 102582859
• Publisher: Elsevier Science
• Year: 1994
• Tongue: English
• Weight: 316 KB
• Volume: 110
• Category: Article
• ISSN: 0021-9991
• DOI: 10.1006/jcph.1994.1029

✦ Synopsis

Ease in finding the configuration at the global energy minimum in a symmetric neural network is important for combinatorial optimization problems. We carry out a comprehensive survey of available strategies for seeking global minima by comparing their performances in the binary representation problem. We recall our previous comparison of steepest descent with analog dynamics, genetic hill-climbing, simulated diffusion, simulated annealing, threshold accepting and simulated tunneling. To this, we add comparisons to other strategies including taboo search and one with field-ordered updating. (C) 1994 Academic Press, tnc.