Figure 9.
Inaccuracy in selected successful fits at different CAR combinations. (A) Incidence of fits with positive parameters only. The 95 CARs were fitted 100 times each, and the number of fits with only positive parameters, the criterion for a successful fit, was counted. The start vector contained the true values. Low CAR numbers generate only a low incidence of successful fits, whereas combinations with ≥12 CARs (cf. Table 1) generated successful fits in all cases. (B) Plot of the mean inaccuracy iaci,sm as a function of the 95 CARs of fits with only positive parameters. The mean was formed first for each successful fit over all parameters xi by Eq. 12a and then over all successful fits specified in A. The inaccuracy decreases toward the highest CAR number of 16.

Inaccuracy in selected successful fits at different CAR combinations. (A) Incidence of fits with positive parameters only. The 95 CARs were fitted 100 times each, and the number of fits with only positive parameters, the criterion for a successful fit, was counted. The start vector contained the true values. Low CAR numbers generate only a low incidence of successful fits, whereas combinations with ≥12 CARs (cf. Table 1) generated successful fits in all cases. (B) Plot of the mean inaccuracy iaci,sm as a function of the 95 CARs of fits with only positive parameters. The mean was formed first for each successful fit over all parameters xi by Eq. 12a and then over all successful fits specified in A. The inaccuracy decreases toward the highest CAR number of 16.

This Feature Is Available To Subscribers Only

or Create an Account

Close Modal
Close Modal