Sökning: onr:"swepub:oai:DiVA.org:kth-302903" >
Convergence of a st...
Convergence of a stochastic gradient method with momentum for non-smooth non-convex optimization
-
- Mai, Vien V. (författare)
- KTH,Reglerteknik
-
- Johansson, Mikael (författare)
- KTH,Reglerteknik
-
(creator_code:org_t)
- International Machine Learning Society (IMLS), 2020
- 2020
- Engelska.
-
Ingår i: 37th International Conference on Machine Learning, ICML 2020. - : International Machine Learning Society (IMLS). ; , s. 6576-6585
- Relaterad länk:
-
https://urn.kb.se/re...
Abstract
Ämnesord
Stäng
- Stochastic gradient methods with momentum are widely used in applications and at the core of optimization subroutines in many popular machine learning libraries. However, their sample complexities have not been obtained for problems beyond those that are convex or smooth. This paper establishes the convergence rate of a stochastic subgradient method with a momentum term of Polyak type for a broad class of non-smooth, non-convex, and constrained optimization problems. Our key innovation is the construction of a special Lyapunov function for which the proven complexity can be achieved without any tuning of the momentum parameter. For smooth problems, we extend the known complexity bound to the constrained case and demonstrate how the unconstrained case can be analyzed under weaker assumptions than the state-of-The-Art. Numerical results confirm our theoretical developments.
Ämnesord
- TEKNIK OCH TEKNOLOGIER -- Elektroteknik och elektronik -- Reglerteknik (hsv//swe)
- ENGINEERING AND TECHNOLOGY -- Electrical Engineering, Electronic Engineering, Information Engineering -- Control Engineering (hsv//eng)
Nyckelord
- Constrained optimization
- Convex optimization
- Gradient methods
- Lyapunov functions
- Machine learning
- Momentum
- Complexity bounds
- Constrained optimi-zation problems
- Convergence rates
- Nonconvex optimization
- Numerical results
- Stochastic gradient methods
- Sub-gradient methods
- Theoretical development
- Stochastic systems
Publikations- och innehållstyp
- ref (ämneskategori)
- kon (ämneskategori)