M
Mattias Fält
Researcher at Lund University
Publications - 20
Citations - 104
Mattias Fält is an academic researcher from Lund University. The author has contributed to research in topics: Convex optimization & Convergence (routing). The author has an hindex of 5, co-authored 20 publications receiving 94 citations.
Papers
More filters
Proceedings ArticleDOI
Line search for averaged operator iteration
TL;DR: In this paper, a line search for averaged iteration is proposed to preserve the theoretical convergence guarantee, while often accelerating practical convergence, and several general cases in which the additional computational cost of the line search is modest compared to the savings obtained.
Posted Content
Line Search for Averaged Operator Iteration
TL;DR: This paper proposes a line search for averaged iteration that preserves the theoretical convergence guarantee, while often accelerating practical convergence, in first order algorithms for convex optimization.
Proceedings ArticleDOI
Optimal convergence rates for generalized alternating projections
Mattias Fält,Pontus Giselsson +1 more
TL;DR: This paper shows how to select the three algorithm parameters to optimize the asymptotic convergence rate, and hence the obtained rate depends on the Friedrichs angle between the subspaces and is considerably better than known rates for other methods such as alternating projections and DouglasRachford splitting.
Proceedings ArticleDOI
Line search for generalized alternating projections
Mattias Fält,Pontus Giselsson +1 more
TL;DR: It is shown that almost all convex optimization problems can be solved using this approach and numerical results show superior performance with both the standard and the projected line search, sometimes by several orders of magnitude, compared to the nominal method.
Journal ArticleDOI
Envelope Functions: Unifications and Further Properties
Pontus Giselsson,Mattias Fält +1 more
TL;DR: In this article, a general envelope function that unifies and generalizes existing envelope functions is presented, which can be used to sharpen corresponding known results for the special cases of forward-backward and Douglas-Rachford splitting.