Parallel minimization algorithms by generalized subdifferentiability
- 10 Downloads
Recently a monotone generalized directional derivative has been introduced for Lipschitz functions. This concept has been applied to represent and optimize nonsmooth functions. The second application resulted relevant for parallel computing, by allowing to define minimization algorithms with high degree of inherent parallelism.
The paper presents first the teoretical background, namely the notions of monotone generalized directional derivative and monotone generalized subdifferential. Then it defines the tools for the procedures, that is a necessary optimality condition and a steepest descent direction. Therefore the minimization algorithms are outlined. Successively the used architectures and the performed numerical experience are described, by listing and commenting the tested functions and the obtained results.
KeywordsLine Search Lipschitz Function Directional Derivative Minimization Algorithm Current Point
Unable to display preview. Download preview PDF.
- Demyanov V.F., Vasilev L.V. (1985), Non differentiable optimization, Opt. Soft. Inc. Publ. Div.Google Scholar
- Rockafellar R.T. (1970), Convex Analysis.Google Scholar
- Sutti C. (1993), On a monotone generalized derivative, Proc. of the Working Day on Math. Optim., Verona 92, Libr. Univ. Ed.Google Scholar
- Sutti C. (to appear), From Monotone Maps to Generalized Convex Functions.Google Scholar