UQLab is designed to have a minimal dependency with MATLAB and its specific toolboxes. UQLab requires at the minimum MATLAB R2014a and the Statistics and Machine Learning Toolbox R2014a (formerly known as the Statistics Toolbox).
To access all the functionalities of UQLab, additional (or newer) toolboxes are required:
- The Optimization Toolbox for the BFGS optimizer used in the Kriging, SVR, and SVC modules and the quadratic programming optimization used in the SVR and SVC modules
- The Global Optimization Toolbox for global optimizers used in the Kriging, SVR, and SVC modules
- The Statistics and Machine Learning Toolbox R2015b for the Sequential Miniminal Optimization (SMO) and Iterative Single Data Algorithm (ISDA) functionalities used in the SVR module
Built-in Optimizers. UQLab is delivered with several global optimizers (e.g., cross-entropy optimizer, CMAES) and allows the Kriging and SVC modules to run even if you don’t have the above-mentioned additional toolboxes.
MATLAB® is a registered trademark of The Mathworks, Inc.