SMA Theory Guide

Most of the computational algorithms implemented in the Overview leverage packages and libraries distributed by other groups such as scikits-learn (wdfTODO REF) and scipy (wdfTODO REF). Therefore, the theory for these algorithms is largely given in the way of links to the original packages. In instances, however, algorithms have been coded within Nodeworks in either pure python or basic numpy calls. This appendix provides additional theory and references for the Nodeworks native algorithms of the Surrogate Modeling and Analysis Toolbox

SMA Theory Guide

• DOE Methods
  • Factorial
  • Covary
  • Monte Carlo
  • Latin Hypercube
  • Central Composite
  • Sobol
  • Hammersly
  • Halton
  • References
• RSM Error Metrics
  • Mean Squared Error
  • Sum of Squared Error
  • R squared
  • L_infinity norm
  • L_1 norm
  • L_2 norm
• Sobol Sensitivity Analysis
  • User Prescribed Parameters

Sensitivity Analysis Methods

TODO aike