The hessian estimation evolution strategy

Publikation: Bidrag til bog/antologi/rapportKonferencebidrag i proceedingsForskningfagfællebedømt

Standard

The hessian estimation evolution strategy. / Glasmachers, Tobias; Krause, Oswin.

Parallel Problem Solving from Nature – PPSN XVI - 16th International Conference, PPSN 2020, Proceedings. red. / Thomas Bäck; Mike Preuss; André Deutz; Michael Emmerich; Hao Wang; Carola Doerr; Heike Trautmann. Springer, 2020. s. 597-609 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), Bind 12269 LNCS).

Publikation: Bidrag til bog/antologi/rapportKonferencebidrag i proceedingsForskningfagfællebedømt

Harvard

Glasmachers, T & Krause, O 2020, The hessian estimation evolution strategy. i T Bäck, M Preuss, A Deutz, M Emmerich, H Wang, C Doerr & H Trautmann (red), Parallel Problem Solving from Nature – PPSN XVI - 16th International Conference, PPSN 2020, Proceedings. Springer, Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), bind 12269 LNCS, s. 597-609, 16th International Conference on Parallel Problem Solving from Nature, PPSN 2020, Leiden, Holland, 05/09/2020. https://doi.org/10.1007/978-3-030-58112-1_41

APA

Glasmachers, T., & Krause, O. (2020). The hessian estimation evolution strategy. I T. Bäck, M. Preuss, A. Deutz, M. Emmerich, H. Wang, C. Doerr, & H. Trautmann (red.), Parallel Problem Solving from Nature – PPSN XVI - 16th International Conference, PPSN 2020, Proceedings (s. 597-609). Springer. Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics) Bind 12269 LNCS https://doi.org/10.1007/978-3-030-58112-1_41

Vancouver

Glasmachers T, Krause O. The hessian estimation evolution strategy. I Bäck T, Preuss M, Deutz A, Emmerich M, Wang H, Doerr C, Trautmann H, red., Parallel Problem Solving from Nature – PPSN XVI - 16th International Conference, PPSN 2020, Proceedings. Springer. 2020. s. 597-609. (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), Bind 12269 LNCS). https://doi.org/10.1007/978-3-030-58112-1_41

Author

Glasmachers, Tobias ; Krause, Oswin. / The hessian estimation evolution strategy. Parallel Problem Solving from Nature – PPSN XVI - 16th International Conference, PPSN 2020, Proceedings. red. / Thomas Bäck ; Mike Preuss ; André Deutz ; Michael Emmerich ; Hao Wang ; Carola Doerr ; Heike Trautmann. Springer, 2020. s. 597-609 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), Bind 12269 LNCS).

Bibtex

@inproceedings{5e668176fbf14f32bd7665f2e57f25de,
title = "The hessian estimation evolution strategy",
abstract = "We present a novel black box optimization algorithm called Hessian Estimation Evolution Strategy. The algorithm updates the covariance matrix of its sampling distribution by directly estimating the curvature of the objective function. This algorithm design is targeted at twice continuously differentiable problems. For this, we extend the cumulative step-size adaptation algorithm of the CMA-ES to mirrored sampling. We demonstrate that our approach to covariance matrix adaptation is efficient by evaluating it on the BBOB/COCO testbed. We also show that the algorithm is surprisingly robust when its core assumption of a twice continuously differentiable objective function is violated. The approach yields a new evolution strategy with competitive performance, and at the same time it also offers an interesting alternative to the usual covariance matrix update mechanism.",
keywords = "Covariance matrix adaptation, Evolution strategy, Hessian matrix",
author = "Tobias Glasmachers and Oswin Krause",
year = "2020",
doi = "10.1007/978-3-030-58112-1_41",
language = "English",
isbn = "9783030581114",
series = "Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)",
publisher = "Springer",
pages = "597--609",
editor = "Thomas B{\"a}ck and Mike Preuss and Andr{\'e} Deutz and Michael Emmerich and Hao Wang and Carola Doerr and Heike Trautmann",
booktitle = "Parallel Problem Solving from Nature – PPSN XVI - 16th International Conference, PPSN 2020, Proceedings",
address = "Switzerland",
note = "16th International Conference on Parallel Problem Solving from Nature, PPSN 2020 ; Conference date: 05-09-2020 Through 09-09-2020",

}

RIS

TY - GEN

T1 - The hessian estimation evolution strategy

AU - Glasmachers, Tobias

AU - Krause, Oswin

PY - 2020

Y1 - 2020

N2 - We present a novel black box optimization algorithm called Hessian Estimation Evolution Strategy. The algorithm updates the covariance matrix of its sampling distribution by directly estimating the curvature of the objective function. This algorithm design is targeted at twice continuously differentiable problems. For this, we extend the cumulative step-size adaptation algorithm of the CMA-ES to mirrored sampling. We demonstrate that our approach to covariance matrix adaptation is efficient by evaluating it on the BBOB/COCO testbed. We also show that the algorithm is surprisingly robust when its core assumption of a twice continuously differentiable objective function is violated. The approach yields a new evolution strategy with competitive performance, and at the same time it also offers an interesting alternative to the usual covariance matrix update mechanism.

AB - We present a novel black box optimization algorithm called Hessian Estimation Evolution Strategy. The algorithm updates the covariance matrix of its sampling distribution by directly estimating the curvature of the objective function. This algorithm design is targeted at twice continuously differentiable problems. For this, we extend the cumulative step-size adaptation algorithm of the CMA-ES to mirrored sampling. We demonstrate that our approach to covariance matrix adaptation is efficient by evaluating it on the BBOB/COCO testbed. We also show that the algorithm is surprisingly robust when its core assumption of a twice continuously differentiable objective function is violated. The approach yields a new evolution strategy with competitive performance, and at the same time it also offers an interesting alternative to the usual covariance matrix update mechanism.

KW - Covariance matrix adaptation

KW - Evolution strategy

KW - Hessian matrix

UR - http://www.scopus.com/inward/record.url?scp=85091292421&partnerID=8YFLogxK

U2 - 10.1007/978-3-030-58112-1_41

DO - 10.1007/978-3-030-58112-1_41

M3 - Article in proceedings

AN - SCOPUS:85091292421

SN - 9783030581114

T3 - Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)

SP - 597

EP - 609

BT - Parallel Problem Solving from Nature – PPSN XVI - 16th International Conference, PPSN 2020, Proceedings

A2 - Bäck, Thomas

A2 - Preuss, Mike

A2 - Deutz, André

A2 - Emmerich, Michael

A2 - Wang, Hao

A2 - Doerr, Carola

A2 - Trautmann, Heike

PB - Springer

T2 - 16th International Conference on Parallel Problem Solving from Nature, PPSN 2020

Y2 - 5 September 2020 through 9 September 2020

ER -

ID: 250555502