A NOVEL NON GRADIENT DEPENDENT METHOD FOR UNCONSTRAINED MULTIVARIATE OPTIMIZATION

Niño Elias D., Posada Hector, Rodriguez Robinson, Toro Luis. A Novel Non Gradient Dependent Method For Unconstrained Multivariate Optimization. Proceedings of the International Conference on Computer and Computational Intelligence, ASME, ISBN: 9780791859926, Bangkok – Thailand, December 2011.

ABSTRACT

This paper states a novel method based on the Hill Climbing for unrestricted multivariate optimization. The proposed method was compared against method from the specialized literature such as Multivariate Newton-Raphson and Multivariate Fletcher-Powell. For making a real comparison, metrics such as Number of Iteration, Processing Time and Stability of the Solution were taken into account. The results showed that the proposed method was the best with a good performance in the metrics, in some cases, of 100% out of 100%.

http://www.asme.org/products/books/international-conference-on-computer-and-computati