Parameter Adjustment
On the previous the least square criteria was described. The data analysis computer program must change the parameter values to achieve a minimum value for the weighted sum of the squared residuals (WSS). This can be illustrated by changing the slope and intercept for the equation for a straight line. The calculated WSS changes with each change in the parameter values.
Figure 6.8.1 Effect on WSS of adjusting slope and intercept
Another more involved example is the calculation of the best fit to data collected after oral administration. Two of the parameters involved in this model are ka and kel. Adjusting the values of these parameters provide different values for the WSS.
Figure 6.8.2 Effect on WSS of adjusting kel and ka
You can download Excel spreadsheets (actually all in the same file) and try your own attempts at reducidng the WSS. Change the parameter values and watch the value for the objective function, WSS, change. With a little 'fiddling' you should be able to get close to a best-fit. Try the straight line example first.
return to the Course index
This page was last modified: Sunday, 28th Jul 2024 at 4:46 pm
Privacy Statement - 25 May 2018
iBook and pdf versions of this material and other PK material is available
Copyright © 2001-2022 David W. A. Bourne (david@boomer.org)