Advertisement

Expanded Options

  • Herbert RueferEmail author
Chapter
  • 442 Downloads

Abstract

Frequently, a prognosis is applied for sets of data composed of many variables. Screening of variables is useful to get rid of those without meaning. This reduces the calculation effort and sometimes improves the accuracy of the prognosis. For some applications, the characteristic outcomes do not exist. For this reason, standardized SNR indices are calculated to take over a similar function. When applying the Mahalanobis algorithm, the squared distance stands for the non-existing numerical outcome and pattern recognition is practicable. Another valuable property of the prognosis is—presupposing that the accuracy of the results is high enough proven by the corresponding SNR index—to analyze existing relationships between variables and the characteristic outcome. According to the remaining parameters (after screening), the adequate size of an orthogonal array is applied. With the assignment of the variables to the orthogonal array and choosing appropriate value settings, all characteristic outcomes are prognosticated and analyzed with the SNR indices. This displays the kind and strength of the contribution of each individual variable to the characteristic outcome without the need of performing any experiment.

Keywords

Orthogonal arrayOrthogonal Array screeningScreening Prognosis Algorithms Inverse Correlation Matrix Integrated Signal Value 
These keywords were added by machine and not by the authors. This process is experimental and the keywords may be updated as the learning algorithm improves.

Copyright information

© Springer Nature Switzerland AG 2019

Authors and Affiliations

  1. 1.BurghausenGermany

Personalised recommendations