Category Archives: Publications

New Book: Multimodal Optimization by Means of Evolutionary Algorithms

Mike Preuss’ book “Multimodal Optimization by Means of Evolutionary Algorithms” is available via Springer. This book

  • Describes state of the art in algorithms, measures and test problems
  • Approaches multimodal optimization algorithms via model-based simulation and statistics
  • Valuable for practitioners with real-world black-box problems

I am looking forward to reading this book.

Just published: Volume 59 – Springer Operations Research/Computer Science Interfaces Series

9781489975461

Here is the announcement from Gabriella Dellino and Carlo Meloni:
We are delighted to announce the publication of the book: “Uncertainty Management in Simulation-Optimization of Complex Systems. Algorithms and Applications” we edited as Volume n. 59 of Springer’s Operations Research/Computer Science Interfaces Series.
The book is available (also in the e-book version) at this link: http://www.springer.com/us/book/9781489975461. Continue reading

Coming Soon: Uncertainty Management in Simulation-Optimization of Complex Systems

The book “Uncertainty Management in Simulation-Optimization of Complex Systems – Algorithms and Applications”, Editors:  Gabriella Dellino and Carlo Meloni is announced on Springer’s Web page. Due date: July 14th, 2015.
It includes the publication “Uncertainty Management Using Sequential Parameter Optimization” (Thomas Bartz-Beielstein, Christian Jung, Martin Zaefferer).

Springer Handbook of Computational Intelligence

They finally arrived this week: Two editor/author copies of the “Springer Handbook of Computational Intelligence“.
The handbook is described by Springer as follows:
“The Springer Handbook for Computational Intelligence is the first book covering the basics, the state-of-the-art and important applications of the dynamic and rapidly expanding discipline of computational intelligence…Content is organized in seven parts:  foundations; fuzzy logic; rough sets; evolutionary computation; neural networks; swarm intelligence and hybrid computational intelligence systems. Each Part is supervised by its own Part Editor(s) so that high-quality content as well as completeness are assured.
Together with Joern Mehnen (Cranfield University) Prof. Bartz-Beielstein served as Part Editor for “Real-World Applications”.

 

Free paper: What Works Best When? A Framework for Systematic Heuristic Evaluation


Ian Dunning, Swapi Gupta, and John Silverholz (Operations Research Center, Massachusetts Institute of Technology) present a systematic review of Max-Cut and Quadratic Unconstrained Binary Optimization (QUBO) heuristics papers. They found only 4% publish source code, only 10% compare heuristics with identical hardware and termination criteria, and most experiments are performed with an artificial, homogeneous set of problem instances.
They state:
Why do we see these limitations in empirical testing? Though best practices for empirical testing have long been published in both the industrial engineering/OR literature (Barr et al. 1995, Rardin and Uzsoy 2001, Silberholz and Golden 2010) and the computer science literature (Johnson 2002, Eiben and Jelasity 2002, Bartz-Beielstein 2006, McGeoch 2012), technical barriers exist to carrying out these best practices.Continue reading

EGO, SPO, SPO+, and more

Steven Ramage from University of British Columbia (UBC) describes the SPO development and the results from the cooperation between UBC and the SPOTSevenLab (Cologne University of Applied Sciences) as follows:

Experimental Methods

Results from the cooperation between UBC and SPOTSeven are published in the book “Experimental Methods for the Analysis of Optimization Algorithms”.

“Bartz-Beielstein et al. [2005] adapted EGO to the optimization of algorithms, with their Sequential Parameter Optimization (SPO) method. SPO uses the same acquisition functions as EGO, but a slightly different model, which includes a second order polynomial fit as well as the standard Gaussian process model. Unlike EGO, their approach is able to deal with random response values through a continual resampling of the best observed points using a doubling strategy, which allows the estimate to converge to the true value over time. Finally, as opposed to fitting the model with each sample point individually, as done by SKO, SPO merges the samples for each point into a better estimate of the objective at that point, and then fits the model on these merged estimates.
Hutter et al. [2009a] directly compared SPO and SKO and their suitability for algorithm configuration. They found that SPO in general outperformed SKO on the algorithms they studied. They also introduced SPO+, which introduced some modifications to the original algorithm….” Continue reading

Energy Frankfurt GRID Report: Make Frankfurt am Main carbon free by 2050

Screen Shot 2015-02-27 at 20.47.42

A proposal to make Frankfurt am Main (Germany) carbon free by 2050 was written by Francisca Molina Moreno, Ivan Leiva Lopez, Gabor Szabo, and Davide Dapelo.

Their idea “is based on a virtual division of the city by areas attending to their production and consumption of energy and also their best option to include one or more energy harvesting system such a PV, Geothermal, Vertical Wind.” To handle timetable demand (dynamic demand)  and production, prediction methods developed in the SPOTSeven Lab [1] are proposed. They write:
“It is therefore known that meteorological information is now available such as quantity of solar radiation, wind and speed direction and also power demand. Statistical methods (stochastic), will provide data source of weather prognosis’ algorithms that will estimate the meteorological parameters [1]. By using technology that can automatically monitor and control the systems in a building by checking the local weather and adjusting climate control for example, energy costs can be reduced by as much as 15%…”

Time series plots of the test data range predictions generated by a method studied in [1].

The technical report for the energy transition in Frankfurt (“Make Frankfurt am Main carbon free by 2050”) is available for download here.

Literature

[1] O. Flasch, M. Friese, K. Vladislavleva, T. Bartz-Beielstein, O. Mersmann, B. Naujoks, J. Stork, and M. Zaefferer. Comparing ensemble-based forecasting methods for smart-metering data. In A. Esparcia- Alc ́azar, editor, Applications of Evolutionary Computation, volume 7835 of Lecture Notes in Computer Science, pages 172–181. Springer Berlin Heidelberg, 2013.  A preprint is available for download from the SPOTSeven page.

[2] Francisca Molina,Ivan Leiva Lopez, Gabor Szabo, and Davide Dapelo. A proposal to make Frankfurt am Main carbon free by 2050. Technical report for the energy transition in Frankfurt. DOI: 10.13140/2.1.1052.2403 Affiliation: Provadis School of International Management and Technology Frankfurt/Main, Germany

Book on Optimization with R(GP)

Paulo Cortez’s new book “Modern Optimization with R”, published by Springer Verlag, contains practical examples on the successful application of RGP, including examples on time series forecasting. RGP models compare favorably to tuned ARIMA models. See http://www.springer.com/mathematics/book/978-3-319-08262-2 for details. There is also an RGP related Webpage: https://rsymbolic.org