Enabling Knowledge Discovery From Simulation-Based Multi-Objective Optimization
Carlos Alberto Barrera Diaz, Henrik Smedberg, Sunith Bandaru & Amos H.C. Ng. Enabling Knowledge Discovery From Simulation-Based...
top of page
Opt-know
OPT-KNOW stands for Optimization ↔ Knowledge. The subject area represents two interconnected lines of research. On one hand explicit human-interpretable knowledge can be extracted from the solutions obtained through multi-objective optimization. This process is called knowledge discovery. On the other hand, the extracted knowledge can in turn be used to improve optimization algorithms. We refer to this process as knowledge-driven optimization or KDO. OPT-KNOW will explore two kinds of KDO, namely online KDO and offline KDO.
Contact:
Sunith Bandaru, Ph.D. Email: sunith.bandaru@his.se. Tel: +46 500 448506
Opt-know in VF-KDO
OPT-KNOW forms the core of KDO in VF-KDO. Industrial optimization problems involve different types of variables, and often specialized techniques may be needed to analyse their solutions. The subject area is responsible for developing new and extending available data mining methods for extracting explicit knowledge from multi-objective optimization data. Additionally, the subject area will also develop improved optimization algorithms that can utilize the extracted knowledge to speed-up the search process or focus search towards regions preferred by the decision maker.
Together with INTERACT and LINK, OPT-KNOW will provide the scientific tools and techniques for knowledge discovery, knowledge visualization, interactive decision support and KDO applications to be undertaken within the other four subject areas.
Opt-know Publications
bottom of page