A Optimal Design

Jump to bottom

Chi-Kuang Yeh edited this page Oct 2, 2017 · 1 revision

Basic Idea

Minimize the -tr(CI) where I is the information matrix under SLSE and C is the direct sum of 0 and Identity matrix Id with dim(Id) = dim(I)-1

To run the program

[del , ANS, error] = D_opt_mod(N,t,theta,range,fun)

Description

Outputs

del ~ fitness value
ANS ~ resulted supporting points and its weights
error ~ maximum measurement error

Parameters

N ~ number of design points
t ~ level of skewness
theta ~ parameter values
range ~ design space

fun ~ function that calculates partial derivatives of the selected model