October 17, 2013

MACROS 1.11.1 Released

DATADVANCE development team is pleased to announce MACROS 1.11.1, the next release in MACROS 1.11 series.

The major feature of this release is the implementation of stochastic variables support in Surrogate Based Optimization (SBO). It enables combining the robust optimization and SBO approaches in the same problem, thus allowing to both account for the uncertainties in problem objectives or constraints and apply surrogate model-based methods to computationally expensive problem functions. Another improvement is the added support for analytical gradients in SBO, which previously was available only in non-SBO problems.

MACROS 1.11.1 also updates the GTApprox technique selection logic: it now includes the Tensor Approximation (TA) and incomplete Tensor Approximation (iTA) techniques, selecting them automatically if the training sample fits TA/iTA technique requirements. The Linear Regression (LR) technique is removed from automatic selection, and in cases when LR was selected previously, the RSM technique will be selected instead.

An important bugfix included in the 1.11.1 release addresses a specific optimization budget violation issue: GTOpt will no longer ignore the budget limit in case when SBO is applied to multi-objective problems (other problem types were not affected). For details on the smaller changes and bugfixes in MACROS 1.11.1, please refer to the release changelog.

Please contact us to receive more information and MACROS updates!