Mathematics > Optimization and Control
[Submitted on 2 Sep 2018 (v1), last revised 3 Feb 2019 (this version, v11)]
Title:The global rate of convergence for optimal tensor methods in smooth convex optimization
View PDFAbstract:We consider convex optimization problems with the objective function having Lipshitz-continuous $p$-th order derivative, where $p\geq 1$. We propose a new tensor method, which closes the gap between the lower $O\left(\varepsilon^{-\frac{2}{3p+1}} \right)$ and upper $O\left(\varepsilon^{-\frac{1}{p+1}} \right)$ iteration complexity bounds for this class of optimization problems.
We also consider uniformly convex functions, and show how the proposed method can be accelerated under this additional assumption. Moreover, we introduce a $p$-th order condition number which naturally arises in the complexity analysis of tensor methods under this assumption.
Finally, we make a numerical study of the proposed optimal method and show that in practice it is faster than the best known accelerated tensor method. We also compare the performance of tensor methods for $p=2$ and $p=3$ and show that the 3rd-order method is superior to the 2nd-order method in practice.
Submission history
From: Eduard Gorbunov [view email][v1] Sun, 2 Sep 2018 20:11:47 UTC (680 KB)
[v2] Thu, 6 Sep 2018 14:19:00 UTC (713 KB)
[v3] Fri, 7 Sep 2018 22:10:48 UTC (713 KB)
[v4] Fri, 21 Sep 2018 20:21:14 UTC (718 KB)
[v5] Sat, 27 Oct 2018 09:47:44 UTC (718 KB)
[v6] Mon, 19 Nov 2018 18:57:25 UTC (718 KB)
[v7] Tue, 20 Nov 2018 16:17:16 UTC (716 KB)
[v8] Sun, 2 Dec 2018 17:16:46 UTC (717 KB)
[v9] Mon, 10 Dec 2018 21:17:07 UTC (715 KB)
[v10] Tue, 25 Dec 2018 13:29:26 UTC (591 KB)
[v11] Sun, 3 Feb 2019 10:55:41 UTC (666 KB)
References & Citations
Bibliographic and Citation Tools
Bibliographic Explorer (What is the Explorer?)
Connected Papers (What is Connected Papers?)
Litmaps (What is Litmaps?)
scite Smart Citations (What are Smart Citations?)
Code, Data and Media Associated with this Article
alphaXiv (What is alphaXiv?)
CatalyzeX Code Finder for Papers (What is CatalyzeX?)
DagsHub (What is DagsHub?)
Gotit.pub (What is GotitPub?)
Hugging Face (What is Huggingface?)
Papers with Code (What is Papers with Code?)
ScienceCast (What is ScienceCast?)
Demos
Recommenders and Search Tools
Influence Flower (What are Influence Flowers?)
CORE Recommender (What is CORE?)
arXivLabs: experimental projects with community collaborators
arXivLabs is a framework that allows collaborators to develop and share new arXiv features directly on our website.
Both individuals and organizations that work with arXivLabs have embraced and accepted our values of openness, community, excellence, and user data privacy. arXiv is committed to these values and only works with partners that adhere to them.
Have an idea for a project that will add value for arXiv's community? Learn more about arXivLabs.