Versatile surrogate-based optimization of medium-scale and large-scale problems.

01 October 2013 → 30 September 2019
Regional and community funding: Special Research Fund, Research Foundation - Flanders (FWO)
Research disciplines
  • Natural sciences
    • Applied mathematics in specific fields
    • Artificial intelligence
    • Computer architecture and networks
    • Distributed computing
    • Information sciences
    • Information systems
    • Programming languages
    • Scientific computing
    • Theoretical computer science
    • Visual computing
    • Other information and computing sciences
    • Other biological sciences
    • Other natural sciences
  • Social sciences
    • Cognitive science and intelligent systems
Gaussian process Kriging optimization
Project description

The main objective of this project is the development of versatile methods based on the Gaussian Process surrogate model to expedite engineering tasks for a wide range of medium-scale (20-100 dimensions) and large-scale (>100 dimensions) problems. The aim is to create a framework of GP-based methods that is easy-to-use (“one-button” approach) and adapts itself to the problem at hand (“self-tuning”) to solve it as efficiently as possible.