Trust-Region Newton-CG with Strong Second-Order Complexity Guarantees for Nonconvex Optimization

Frank Curtis (1) , Daniel Robinson (1) , Clément Royer (2, 3) , Stephen Wright (4)

1 ISE - Department of Industrial and Systems Engineering [Lehigh]
2 LAMSADE - Laboratoire d'analyse et modélisation de systèmes pour l'aide à la décision
3 MILES - Machine Intelligence and Learning Systems
4 University of Wisconsin-Madison

Domaines: Optimisation et contrôle [math.OC]

Frank Curtis, Daniel Robinson, Clément Royer, Stephen Wright. Trust-Region Newton-CG with Strong Second-Order Complexity Guarantees for Nonconvex Optimization. SIAM Journal on Optimization, 2021, 31 (1), pp.518-544. ⟨10.1137/19M130563X⟩. ⟨hal-03135526⟩