Neural-network-assisted genetic algorithm applied to silicon clusters

Journal Article · Sat Mar 01 04:00:00 EST 2003 · Physical Review. A

DOI:https://doi.org/10.1103/PhysRevA.67.033203· OSTI ID:20633856

Marim, L R; Lemes, M R; Pino, A Jr dal ^[1]

---

1. Department of Physics, Instituto Tecnologico de Aeronautica, Pca. Marechal Eduardo Gomes, 50-Sao Jose dos Campos, Sao Paulo 12228-900 (Brazil)

Recently, a new optimization procedure that combines the power of artificial neural-networks with the versatility of the genetic algorithm (GA) was introduced. This method, called neural-network-assisted genetic algorithm (NAGA), uses a neural network to restrict the search space and it is expected to speed up the solution of global optimization problems if some previous information is available. In this paper, we have tested NAGA to determine the ground-state geometry of Si{sub n} (10{<=}n{<=}15) according to a tight-binding total-energy method. Our results indicate that NAGA was able to find the desired global minimum of the potential energy for all the test cases and it was at least ten times faster than pure genetic algorithm.

• OSTI ID: 20633856
• Journal Information: Physical Review. A, Journal Name: Physical Review. A Journal Issue: 3 Vol. 67; ISSN 1050-2947; ISSN PLRAAN
• Country of Publication: United States
• Language: English

Related Subjects

74 ATOMIC AND MOLECULAR PHYSICS
ALGORITHMS
ATOMIC CLUSTERS
GEOMETRY
GROUND STATES
NEURAL NETWORKS
OPTIMIZATION
POTENTIAL ENERGY
SILICON