Understanding of optimization algorithm’s behavior is a vital part that is needed for quality progress in the field of stochastic optimization algorithms. Too often (new) algorithms are setup and tuned only focusing on achieving the desired optimization goal. While this might be effective and efficient in short term, in long term this is insufficient due to the fact that this needs to be repeated for every new problem that arises. Such approach provides only minor immediate gains, instead of contributing to the progress in research on optimization algorithms. To be able to overcome this deficiency, we need to establish new standards for understanding optimization algorithm behavior, which will provide understanding of the working principles behind the stochastic optimization algorithms. This includes theoretical and empirical research, which would lead to providing insight into answering questions such as (1) why does an algorithm work for some problems but does not work for others, (2) how to explore the fitness landscape to gain better understanding of the algorithm’s behavior, and (3) how to interconnect stochastic optimization and machine learning to improve the algorithm’s behavior on new unseen instances.
The focus of this workshop is to highlight theoretical and empirical research that investigate approaches needed to analyze stochastic optimization algorithms and performance assessment with regard to different criteria. The main goal is to bring the problem and importance of understanding optimization algorithms closer to researchers and to show them how and why this is important for future development in the optimization community. This will help researchers/users to transfer the gained knowledge from theory into the real world, or to find the algorithm that is best suited to the characteristics of a given real-world problem.
Faculty of Science
University of Nottingham
Title: Towards a Better Understanding of Search Algorithms: A New Standard on Algorithm Design
Along with the recent successes of machine learning to numerous applications comes the next challenge of general AI optimisation algorithms. In research, intelligent search algorithms are often designed case by case for specific problems. The rich knowledge of algorithm design is scattered and often discarded, leading to huge waste. In practice, practitioners are faced with the barrier of extensive expertise required to design effective algorithms. There is a lack of standard in algorithm design to support a deeper understanding on and retain coherent knowledge of good behaviours of search algorithms. Such standard is key to develop effective algorithms in both research and practice efficiently.
This talk introduces a new standard on algorithm design based on algorithmic components, and discuss some recent results on automated algorithm design for combinatorial optimisation problems. With the new standard, the new algorithms automatically evolved can be modelled in a consistent form of algorithmic components configured in the best way. These configurations thus can potentially provide a way to gain better understanding of algorithm behaviours. The talk will also present research on the search spaces of hyper-heuristics. As one of the general search algorithms, hyper-heuristics can also be modelled using the new standard, findings on which provide insights for future research on the new standard addressing various optimisation problems.
All submissions should reflect the CEC2019 submission format provided at http://www.cec2019.org/papers.html#submission, will be handled through Easychair (https://easychair.org/conferences/?conf=ueob2019) and reviewed by the program committee.
In order to participate to this workshop, full or student registration of CEC 2019 is needed.
Selected papers will be invited to be extended for a special issue in the Natural Computing.
Department of Biomedical Data Sciences, Center for Population Health Sciences, Stanford Medicine
Stanford University, USA
Computer Systems Depratment
Jožef Stefan Institue, Slovenia
Artificial Intelligence Research Institute (IIIA)
Spanish National Research Council (CSIC), Spain