Version 1.1.4

Minor Changes:

• More flixibility for visualizing the Pareto inputs
• Debugging GUI

Version 1.1.3

MCMC can now be used for finding a potential optimum based on the Kriging prediction

New Feature: Multi-objective global optimization (MOGO)

Process optimization is typically complicated by conflicting design targets, such as productivity and yield. Multi-Objective Global Optimization (MOGO) algorithm utilizes statistical regression models for multi-objective optimization. Kriging is used for estimating functional relationships between design variables and performance indicators . The Pareto front is iteratively approximated by planning new experiments such as to maximize the Expected Hypervolume Improvement (EHVI) as determined from the Kriging models by Markov Chain Monte Carlo (MCMC) sampling.

Version 1.1.0

Major Changes:

• Inequality Constraint w.r.t. to objectives can now be applied for expected improvement
• Modification in plotParetoFront and predictParetoFront regarding the 2D case
• Debugging of AnalyzeMixture Class
• Heterodedastic noise can now be estimated (generateNoiseModel)
• Major improvements in TaylorKriging (Kriging with non-linear mechanistic models)
• Speedup in calculation of 2D expected hypervolume improvement

Minor Changes:

• Small changes w.r.t. to the tutorial
• New color scheme available: FZJ
• Use now "surf" instead of "mesh" for 3D plots
• The class BaysianOptimization was renamed to BaysianOptimizationClass

Version 1.0.4

Bug fix:

• Some minor bug fixes in the Anlysis and Super class
• Plotting optimal regions is now enabled for Screening plots

Version 1.0.3

Bug fix:

• Some minor bug fixes of the KrigingGUI
• Basis function can now also be displyed in case of 3D interpolation

Version 1.0.2

Major improvements since last release:

• Contour plots can now be created for 3D interpolation

Bug fix:

• Measurement noise is now considered in conditional simulation

Release Version

20160701

- Editing Cemments in "doANOVAandPrediction", "doCompositeDesignAnalysis", and "doFullFactorialAnalysis"
- Enable "plotContour" for 3D Interpolation
-> Editing "labelPlots_23D" and "plotInterpolation_23D"
- Debug doConditionalSimulation_ResAlgInEq
-> Now also measurment noise is considered


Signed-off-by: Lazloo <[email protected]>