agaier / sail_gecco2017

Code used to produce results in "Data-Efficient Exploration, Optimization, and Modeling of Diverse Designs through Surrogate-Assisted Illumination (SAIL)", presented at GECCO 2017. Posted in the interest of replicability and so will not be updated. General use production code is at 'github.com/agaier/sail'. Stay tuned for 'sferes2' C++ code.

sail_gecco2017

source code for experiments in "Data-Efficient Exploration, Optimization, and Modeling of Diverse Designs through Surrogate-Assisted Illumination" presented at GECCO 2017.

https://scholar.google.com/scholar?cluster=5948437257089311286&hl=en&as_sdt=0,5

'help sail' for contents of source code and main algorithm syntax

Produced using Matlab Version: 9.1.0.441655 (R2016b)

Required Toolboxes: Parallel Computing (for precise evaluation speed up) Statistics and Machine Learning (for Sobol sequence)

Required Software: XFOIL low Rn airfoil design and analysis code (raphael.mit.edu/xfoil/)

Includes: CMAES Version 3.61.beta, Hansen (help cmaes) GMPL Version 3.6, Rasmussen & Nickisch (help gpml)