Open problem: Fast stochastic exp-concave optimization

Tomer Koren

Open problem: Fast stochastic exp-concave optimization

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution › peer-review

Abstract

Stochastic exp-concave optimization is an important primitive in machine learning that captures several fundamental problems, including linear regression, logistic regression and more. The exp-concavity property allows for fast convergence rates, as compared to general stochastic optimization. However, current algorithms that attain such rates scale poorly with the dimension n and run in time O(n⁴), even on very simple instances of the problem. The question we pose is whether it is possible to obtain fast rates for exp-concave functions using more computationally-efficient algorithms.

Original language	English
Title of host publication	Proceedings of the 26th Annual Conference on Learning Theory
Editors	Shai Shalev-Shwartz, Ingo Steinwart
Pages	1073-1075
Number of pages	3
State	Published - 2013
Externally published	Yes
Event	26th Conference on Learning Theory, COLT 2013 - Princeton, NJ, United States Duration: 12 Jun 2013 → 14 Jun 2013

Publication series

Name	Proceedings of Machine Learning Research
Publisher	PMLR
Volume	30

Conference

Conference	26th Conference on Learning Theory, COLT 2013
Country/Territory	United States
City	Princeton, NJ
Period	12/06/13 → 14/06/13

All Science Journal Classification (ASJC) codes

Software
Control and Systems Engineering
Statistics and Probability
Artificial Intelligence

Access to Document

https://proceedings.mlr.press/v30/Koren13.html

Cite this

@inproceedings{a991036037e1441caaf7097a235db005,

title = "Open problem: Fast stochastic exp-concave optimization",

abstract = "Stochastic exp-concave optimization is an important primitive in machine learning that captures several fundamental problems, including linear regression, logistic regression and more. The exp-concavity property allows for fast convergence rates, as compared to general stochastic optimization. However, current algorithms that attain such rates scale poorly with the dimension n and run in time O(n4), even on very simple instances of the problem. The question we pose is whether it is possible to obtain fast rates for exp-concave functions using more computationally-efficient algorithms.",

author = "Tomer Koren",

year = "2013",

language = "الإنجليزيّة",

series = "Proceedings of Machine Learning Research",

publisher = "PMLR",

pages = "1073--1075",

editor = "Shai Shalev-Shwartz and Ingo Steinwart",

booktitle = "Proceedings of the 26th Annual Conference on Learning Theory",

note = "26th Conference on Learning Theory, COLT 2013 ; Conference date: 12-06-2013 Through 14-06-2013",

}

Koren, T 2013, Open problem: Fast stochastic exp-concave optimization. in S Shalev-Shwartz & I Steinwart (eds), Proceedings of the 26th Annual Conference on Learning Theory. Proceedings of Machine Learning Research, vol. 30, pp. 1073-1075, 26th Conference on Learning Theory, COLT 2013, Princeton, NJ, United States, 12/06/13. <https://proceedings.mlr.press/v30/Koren13.html>

TY - GEN

T1 - Open problem

T2 - 26th Conference on Learning Theory, COLT 2013

AU - Koren, Tomer

PY - 2013

Y1 - 2013

N2 - Stochastic exp-concave optimization is an important primitive in machine learning that captures several fundamental problems, including linear regression, logistic regression and more. The exp-concavity property allows for fast convergence rates, as compared to general stochastic optimization. However, current algorithms that attain such rates scale poorly with the dimension n and run in time O(n4), even on very simple instances of the problem. The question we pose is whether it is possible to obtain fast rates for exp-concave functions using more computationally-efficient algorithms.

AB - Stochastic exp-concave optimization is an important primitive in machine learning that captures several fundamental problems, including linear regression, logistic regression and more. The exp-concavity property allows for fast convergence rates, as compared to general stochastic optimization. However, current algorithms that attain such rates scale poorly with the dimension n and run in time O(n4), even on very simple instances of the problem. The question we pose is whether it is possible to obtain fast rates for exp-concave functions using more computationally-efficient algorithms.

UR - http://www.scopus.com/inward/record.url?scp=84898037492&partnerID=8YFLogxK

M3 - منشور من مؤتمر

T3 - Proceedings of Machine Learning Research

SP - 1073

EP - 1075

BT - Proceedings of the 26th Annual Conference on Learning Theory

A2 - Shalev-Shwartz, Shai

A2 - Steinwart, Ingo

Y2 - 12 June 2013 through 14 June 2013

ER -

Open problem: Fast stochastic exp-concave optimization

Abstract

Publication series

Conference

All Science Journal Classification (ASJC) codes

Access to Document

Other files and links

Fingerprint

Cite this