Eikos - Sensitivity Analysis Matlab Toolbox

Eikos is a toolbox that supports “state of the art” sensitivity analysis methods (local as well as global). Sensitivity analysis (SA) is used to assess the influence of model parameters on model predictions.

The toolbox is developed in MATLAB^® and is therefore best suited for models developed under the Matlab/Simulink^® environment, although it can also be used together with stand-alone model codes developed in other programming languages.

Distributions: (Log)-Uniform, (Log)-Triangular, (Log)-Normal, Weibull, Exponential, Gamma and Beta.

Correlations between parameters may be induced by rank order correlation (method of Iman and Conover). Sampling techniques: Simple random, Latin Hypercube, Quasi-random LPt.

Included global SA methods:

• Standardized (Ranked) Regression Coefficients
• Pearson Product Moment Correlation Coefficients
• Spearman Ranked Correlation Coefficients
• Partial (Ranked) Correlation Coefficients
• Smirnov and Cramer von Misès two-sample tests
• Morris's OAT designs
• Method of Sobol'
• Extended Fourier Amplitude Sensitivity Test (EFAST) and the classical FAST method

Eikos provides a Graphical User Interface (GUI) which let you:

• specify uncertain model parameters and model outputs of interest
• assign probability distribution functions to describe parameter uncertainty
• perform Sensitivity Analysis of model predictions
• perform Uncertainty Analysis of model predictions
• graphically review the results using pie charts, scatter plots, tornado graphs, etc.
• import/export individual assessments
• export results to report charts in text or excel format

Publication:

• Eikos: A Simulation Toolbox for Sensitivity Analysis *
• Poster (Proceedings of The 2^nd International Conference on Radioactivity in the Environment, ISBN 82-90362-21-8. 2-6 October 2005 in Nice, France. 361-364.) *

* Requires Adobe Reader