Category Archives: Optimization

Paper “Optimization and Adaptation of a Resource Planning Tool for Hospitals Under Special Consideration of the COVID-19 Pandemic” accepted for oral presentation at the IEEE CEC 2021

Authors:
Thomas Bartz-Beielstein, Marcel Droescher, Alpar Guer, Alexander Hinterleitner, Tom Lawton, Olaf Mersmann, Dessislava Peeva, Lennard Reese, Frederik Rehbach, Nicolas Rehbach, Amrita Sen, Aleksandr Subbotin and Martin Zaefferer

Abstract:
Hospitals and health-care institutions need to plan the resources required for
handling the increased load, i.e., beds and ventilators during the COVID-19
pandemic.
BaBSim.Hospital, an open-source tool for capacity planning based on discrete
event simulation, was developed over the last year to support doctors,
administrations, health authorities, and crisis teams in Germany.
To obtain reliable results, 29 simulation parameters such as durations and
probabilities must be specified. While reasonable default values were obtained
in detailed discussions with medical professionals, the parameters have to be
regularly and automatically optimized based on current data.

We aim to investigate how a set of parameters that is tailored to the German
health system can be transferred to other regions. Therefore, we will use data
from the UK. Our study demonstrates the flexibility of the discrete event
simulation approach. However, transferring the optimal German parameter settings
to the UK situation does not work—parameter ranges must be modified. The
adaptation has been shown to reduce simulation errors by nearly 70%.
The simulation-via-optimization approach is not restricted to health-care
institutions, it is applicable to many other real-world problems, e.g., the
development of new elevator systems to cover the last mile or simulation of
student flow in academic study periods.

Link to the simulator:
https://covid-resource-sim.th-koeln.de/app/babsim.hospitalvis

Link to the open-source software:
https://cran.r-project.org/package=babsim.hospital

Update: GECCO 2021 Industrial Challenge – Competition on Optimization of a Simulation Model for Capacity and Resource Planing Task for hospitals under special consideration of the COVID-19 Pandemic

Competition website:https://www.th-koeln.de/informatik-und-ingenieurwissenschaften/gecco-2021-industrial-challenge-call-for-participation_82086.php

Highlights of the GECCO 2021 Industrial Challenge include:

  • Interesting Problem Domain: The impact of COVID-19 on the health system is ongoing and tools to help in capacity planning are more important than ever.
  • Real-world Problems: Test your algorithms and methods, directly on real and current data.
  • Easy Access: Easily Participate through our online platform, no installations required.
  • Fair Submission Assessment: Winners are determined automatically through our online portal, fully objectively, only based on the final result quality.
  • Publication Possibilities: We are able to accept 2-page submissions for the GECCO Companion; thus, publications are possible directly through competition participation.
  • Price money:  The best solution will receive 300€, the second place 200€ and the third place 100€

Simulation models are valuable tools for resource usage estimation and capacity planning. The simulator, BaBSim.Hospital, explicitly covers difficulties for hospitals caused by the COVID-19 pandemic. The simulator can handle many aspects of resource planning in hospitals, such as ICU beds, ventilators or personal, while taking into consideration several cohorts as age or current health status. 
The task represents an instance of an expensive, high-dimensional computer simulation-based optimization problem. The simulations will be executed through an interface and hosted on one of our servers (similar to our last year’s challenge).

Your goal is to find an optimal parameter configuration for the BaBSim.Hospital simulator with a very limited budget of objective function evaluations. The participants will be free to apply one or multiple optimization algorithms of their choice.

Like last year, we are able to provide the opportunity for all participants to submit 2-page algorithm descriptions for the GECCO Companion. Thus, it is now possible to create publications in a similar procedure to the Late-Breaking Abstracts (LBAs) directly through competition participation!

2-Page Algorithm Description Submission Deadline: 2021-04-12 23:59
To be held as part of the 2021 Genetic and Evolutionary Computation Conference (GECCO 2021) organized by ACM SIGEVO (https://gecco-2021.sigevo.org).

Tuning algorithms for black-box optimization: State of the art and future perspectives will be available in May 2021

The publication “Tuning algorithms for black-box optimization: State of the art and future perspectives” by Thomas Bartz-Beielstein, Frederik Rehbach, and Margarita Rebolledo will be published as a contribution to the book Black Box Optimization, Machine Learning and No- Free Lunch Theorems. The book is edited by Panos Pardalos, Varvara Rasskazova, and Michael Vrahatis and will be number 170 in Springer Optimization and Its Applications series, see https://www.springer.com/de/book/9783030665142

Continue reading

“Sequential Parameter Optimization for Mixed-Discrete Problems” Published in the Book “Optimization Under Uncertainty with Applications to Aerospace Engineering”

The book Optimization Under Uncertainty with Applications to Aerospace Engineering has recently been published in electronic and print format. It includes our contribution Sequential Parameter Optimization for Mixed-Discrete Problems (Lorenzo Gentile, Thomas Bartz-Beielstein, and Martin Zaefferer).

GECCO Challenge 2021 is Online

The GECCO Industrial Challenge is a yearly competition for CI researchers and practitioners taking place at the GECCO. It poses difficult real-world problems provided by industry partners from various fields.

Simulation models are valuable tools for resource usage estimation and capacity planning. The simulator, BaBSim.Hospital, explicitly covers difficulties for hospitals caused by the COVID-19 pandemic. The simulator can handle many aspects of resource planning in hospitals, such as ICU beds, ventilators or personal, while taking into consideration several cohorts as age or current health status.

The task represents an instance of an expensive, high-dimensional computer simulation-based optimization problem. The simulations will be executed through an interface and hosted on one of our servers (similar to our last year’s challenge).

Your goal is to find an optimal parameter configuration for the BaBSim.Hospital simulator with a very limited budget of objective function evaluations. The participants will be free to apply one or multiple optimization algorithms of their choice.

More information can be found here.

Free Preprint: “Expected Improvement versus Predicted Value in Surrogate-Based #Optimization” Available on Cologne Open Science

The publication “Expected Improvement versus Predicted Value in Surrogate-Based Optimization”, written by Frederik Rehbach, Martin Zaefferer, Boris Naujoks, and Thomas Bartz-Beielstein, deals with the correct parameterization for model-based optimization algorithms. The publication is available on Cologne Open Science.

Continue reading

New on arXiv: “Benchmarking in Optimization: Best Practice and Open Issues”

The most recent version of the article “Benchmarking in Optimization: Best Practice and Open Issues”, which was written by Thomas Bartz-Beielstein, Carola Doerr, Jakob Bossek, Sowmya Chandrasekaran, Tome Eftimov, Andreas Fischbach, Pascal Kerschke, Manuel Lopez-Ibanez, Katherine M. Malan, Jason H. Moore, Boris Naujoks, Patryk Orzechowski, Vanessa Volz, Markus Wagner, and Thomas Weise can be downloaded from arXiv http://arxiv.org/abs/2007.03488.

Survey “Benchmarking in Optimization: Best Practice and Open Issues” available

This survey compiles ideas and recommendations from more than a dozen
researchers with different backgrounds and from different institutes around the
world. Promoting best practice in benchmarking is its main goal. The article
discusses eight essential topics in benchmarking: clearly stated goals, well-
specified problems, suitable algorithms, adequate performance measures,
thoughtful analysis, effective and efficient designs, comprehensible
presentations, and guaranteed reproducibility. The final goal is to provide
well-accepted guidelines (rules) that might be useful for authors and
reviewers. As benchmarking in optimization is an active and evolving field of
research this manuscript is meant to co-evolve over time by means of periodic
updates.

The PDF version of this survey is available here and will be published on arXiv soon.