---
_id: '47522'
abstract:
- lang: eng
text: Artificial benchmark functions are commonly used in optimization research
because of their ability to rapidly evaluate potential solutions, making them
a preferred substitute for real-world problems. However, these benchmark functions
have faced criticism for their limited resemblance to real-world problems. In
response, recent research has focused on automatically generating new benchmark
functions for areas where established test suites are inadequate. These approaches
have limitations, such as the difficulty of generating new benchmark functions
that exhibit exploratory landscape analysis (ELA) features beyond those of existing
benchmarks.The objective of this work is to develop a method for generating benchmark
functions for single-objective continuous optimization with user-specified structural
properties. Specifically, we aim to demonstrate a proof of concept for a method
that uses an ELA feature vector to specify these properties in advance. To achieve
this, we begin by generating a random sample of decision space variables and objective
values. We then adjust the objective values using CMA-ES until the corresponding
features of our new problem match the predefined ELA features within a specified
threshold. By iteratively transforming the landscape in this way, we ensure that
the resulting function exhibits the desired properties. To create the final function,
we use the resulting point cloud as training data for a simple neural network
that produces a function exhibiting the target ELA features. We demonstrate the
effectiveness of this approach by replicating the existing functions of the well-known
BBOB suite and creating new functions with ELA feature values that are not present
in BBOB.
author:
- first_name: Raphael Patrick
full_name: Prager, Raphael Patrick
last_name: Prager
- first_name: Konstantin
full_name: Dietrich, Konstantin
last_name: Dietrich
- first_name: Lennart
full_name: Schneider, Lennart
last_name: Schneider
- first_name: Lennart
full_name: Schäpermeier, Lennart
last_name: Schäpermeier
- first_name: Bernd
full_name: Bischl, Bernd
last_name: Bischl
- first_name: Pascal
full_name: Kerschke, Pascal
last_name: Kerschke
- first_name: Heike
full_name: Trautmann, Heike
id: '100740'
last_name: Trautmann
orcid: 0000-0002-9788-8282
- first_name: Olaf
full_name: Mersmann, Olaf
last_name: Mersmann
citation:
ama: 'Prager RP, Dietrich K, Schneider L, et al. Neural Networks as Black-Box Benchmark
Functions Optimized for Exploratory Landscape Features. In: Proceedings of
the 17th ACM/SIGEVO Conference on Foundations of Genetic Algorithms. FOGA
’23. Association for Computing Machinery; 2023:129–139. doi:10.1145/3594805.3607136'
apa: Prager, R. P., Dietrich, K., Schneider, L., Schäpermeier, L., Bischl, B., Kerschke,
P., Trautmann, H., & Mersmann, O. (2023). Neural Networks as Black-Box Benchmark
Functions Optimized for Exploratory Landscape Features. Proceedings of the
17th ACM/SIGEVO Conference on Foundations of Genetic Algorithms, 129–139.
https://doi.org/10.1145/3594805.3607136
bibtex: '@inproceedings{Prager_Dietrich_Schneider_Schäpermeier_Bischl_Kerschke_Trautmann_Mersmann_2023,
place={New York, NY, USA}, series={FOGA ’23}, title={Neural Networks as Black-Box
Benchmark Functions Optimized for Exploratory Landscape Features}, DOI={10.1145/3594805.3607136},
booktitle={Proceedings of the 17th ACM/SIGEVO Conference on Foundations of Genetic
Algorithms}, publisher={Association for Computing Machinery}, author={Prager,
Raphael Patrick and Dietrich, Konstantin and Schneider, Lennart and Schäpermeier,
Lennart and Bischl, Bernd and Kerschke, Pascal and Trautmann, Heike and Mersmann,
Olaf}, year={2023}, pages={129–139}, collection={FOGA ’23} }'
chicago: 'Prager, Raphael Patrick, Konstantin Dietrich, Lennart Schneider, Lennart
Schäpermeier, Bernd Bischl, Pascal Kerschke, Heike Trautmann, and Olaf Mersmann.
“Neural Networks as Black-Box Benchmark Functions Optimized for Exploratory Landscape
Features.” In Proceedings of the 17th ACM/SIGEVO Conference on Foundations
of Genetic Algorithms, 129–139. FOGA ’23. New York, NY, USA: Association for
Computing Machinery, 2023. https://doi.org/10.1145/3594805.3607136.'
ieee: 'R. P. Prager et al., “Neural Networks as Black-Box Benchmark Functions
Optimized for Exploratory Landscape Features,” in Proceedings of the 17th ACM/SIGEVO
Conference on Foundations of Genetic Algorithms, 2023, pp. 129–139, doi: 10.1145/3594805.3607136.'
mla: Prager, Raphael Patrick, et al. “Neural Networks as Black-Box Benchmark Functions
Optimized for Exploratory Landscape Features.” Proceedings of the 17th ACM/SIGEVO
Conference on Foundations of Genetic Algorithms, Association for Computing
Machinery, 2023, pp. 129–139, doi:10.1145/3594805.3607136.
short: 'R.P. Prager, K. Dietrich, L. Schneider, L. Schäpermeier, B. Bischl, P. Kerschke,
H. Trautmann, O. Mersmann, in: Proceedings of the 17th ACM/SIGEVO Conference on
Foundations of Genetic Algorithms, Association for Computing Machinery, New York,
NY, USA, 2023, pp. 129–139.'
date_created: 2023-09-27T15:43:17Z
date_updated: 2023-10-16T12:33:02Z
department:
- _id: '34'
- _id: '819'
doi: 10.1145/3594805.3607136
keyword:
- Benchmarking
- Instance Generator
- Black-Box Continuous Optimization
- Exploratory Landscape Analysis
- Neural Networks
language:
- iso: eng
page: 129–139
place: New York, NY, USA
publication: Proceedings of the 17th ACM/SIGEVO Conference on Foundations of Genetic
Algorithms
publication_identifier:
isbn:
- '9798400702020'
publisher: Association for Computing Machinery
series_title: FOGA ’23
status: public
title: Neural Networks as Black-Box Benchmark Functions Optimized for Exploratory
Landscape Features
type: conference
user_id: '15504'
year: '2023'
...