# zbMATH — the first resource for mathematics

Complexity of gradient descent for multiobjective optimization. (English) Zbl 1429.90067
Summary: A number of first-order methods have been proposed for smooth multiobjective optimization for which some form of convergence to first-order criticality has been proved. Such convergence is global in the sense of being independent of the starting point. In this paper, we analyse the rate of convergence of gradient descent for smooth unconstrained multiobjective optimization, and we do it for non-convex, convex, and strongly convex vector functions. These global rates are shown to be the same as for gradient descent in single-objective optimization and correspond to appropriate worst-case complexity bounds. In the convex cases, the rates are given for implicit scalarizations of the problem vector function.

##### MSC:
 90C29 Multi-objective and goal programming 90C60 Abstract computational complexity for mathematical programming problems
Full Text:
##### References:
 [1] Attouch, H.; Goudou, X., A continuous gradient-like dynamical approach to pareto-optimization in Hilbert spaces, Set-Valued. Var. Anal., 22, 189-219, (2014) · Zbl 1317.90261 [2] Beck, A., First-order Methods in Optimization, (2017), SIAM: SIAM, Philadelphia · Zbl 1384.65033 [3] Birgin, E. G.; Gardenghi, J. L.; Martínez, J. M.; Santos, S. A.; Toint, Ph. L., Worst-case evaluation complexity for unconstrained nonlinear optimization using high-order regularized models, Math. Program., 163, 359-368, (2017) · Zbl 1365.90236 [4] Bonnel, H.; Iusem, A. N.; Svaiter, B. F., Proximal methods in vector optimization, SIAM J. Optim., 15, 953-970, (2005) · Zbl 1093.90054 [5] Carrizo, G. A.; Lotito, P. A.; Maciel, M. C., Trust region globalization strategy for the nonconvex unconstrained multiobjective optimization problem, Math. Program., 159, 339-369, (2016) · Zbl 1345.90081 [6] Cartis, C.; Gould, N. I.M.; Toint, Ph. L., On the complexity of steepest descent, Newton’s and regularized Newton’s methods for nonconvex unconstrained optimization, SIAM J. Optim., 20, 2833-2852, (2010) · Zbl 1211.90225 [7] Fliege, J.; Graña Drummond, L. M.; Svaiter, B. F., Newton’s method for multiobjective optimization, SIAM J. Optim., 20, 602-626, (2009) · Zbl 1195.90078 [8] Fliege, J.; Svaiter, B. F., Steepest descent methods for multicriteria optimization, Math. Methods Oper. Res., 51, 479-494, (2000) · Zbl 1054.90067 [9] Fukuda, E. H.; Graña Drummond, L. M., A survey on multiobjective descent methods, Pesquisa Operacional, 34, 585-620, (2014) [10] Três Contribuições em Otimização Não Linear e Não Convexa, Ph.D. thesis, Universidade Federal do Paraná, 2014. In Portuguese [11] On the worst-case complexity of projected gradient methods for convex constrained multiobjective optimization. 2016 [12] Grapiglia, G. N.; Yuan, J.; Yuan, Y.-X., On the convergence and worst-case complexity of trust-region and regularization methods for unconstrained optimization, Math. Program., 152, 491-520, (2015) · Zbl 1319.90065 [13] Jahn, J., Vector Optimization, (2009), Springer: Springer, Berlin · Zbl 0577.90078 [14] Non-linear conjugate gradient methods for vector optimization, preprint (2017), Federal University of Goias. [15] Nesterov, Y., A method of solving a convex programming problem with convergence rate $$####,$$ Soviet Math. Doklady, 27, 372-376, (1983) · Zbl 0535.90071 [16] Nesterov, Y., Introductory Lectures on Convex Optimization, (2004), Kluwer Academic Publishers: Kluwer Academic Publishers, Dordrecht · Zbl 1086.90045 [17] Vicente, L. N., Worst case complexity of direct search, EURO J. Comput. Optim., 1, 143-153, (2013) · Zbl 1304.90198
This reference list is based on information provided by the publisher or from digital mathematics libraries. Its items are heuristically matched to zbMATH identifiers and may contain data conversion errors. It attempts to reflect the references listed in the original paper as accurately as possible without claiming the completeness or perfect precision of the matching.