Skip to main content

Table 2 Fit of Training Data to Estimating Functions

From: Estimation of the number of extreme pathways for metabolic networks

 

Fit of Factors

Over-Estimation

Under-Estimation

Factor

Correlation r

m.a.e.

r.m.s.

Max. Value

Members>1

Min. Value

Members<1

θ1 = ∑d+d-c

0.883439

0.407565

0.539834

1.512650

5

-1.138180

1

θ1 = ∑d+d-

0.887057

0.406882

0.544633

1.585230

3

-1.071100

1

R ω θ 1

0.900704

0.383868

0.512515

1.629060

3

-0.918259

0

R ω θ 1

0.898332

0.381380

0.518272

1.674300

3

-0.963172

0

  1. This table summarizes the statistics describing the relationships between the four different estimations f i (θ j ) and the 52 data points. The results show that the two factors θ1 and θ2 allowed better estimations when the scaling factor Rωwas included. The top two rows and the bottom two rows of the table, respectively, represent the factors before and after scaling by the factor Rω. The cells with the highest Pearson's correlation, r, before and after scaling are emphasized with † and ‡, respectively.