Table 2. The parameters of the linear regression equations obtained during learnig and test results.
|
Learning set |
R2 L |
MSEL |
Q2 |
N |
MSELOO |
R2T |
SET |
R2T (set 3) |
SET (set 3) |
|
SET1 |
0.73 |
0.43 |
0.40 |
9 |
0.72 |
0.04 |
0.70 |
0.14 |
2.56 |
|
SET1 |
0.72 |
0.43 |
0.48 |
6 |
0.72 |
0.04 |
0.70 |
0.15 |
2.48 |
|
SET2 |
0.16 |
0.48 |
0.03 |
9 |
0.74 |
0.08 |
0.03 |
0.03 |
0.36 |
|
SET1+SET2 |
0.47 |
0.59 |
0.26 |
9 |
0.72 |
- |
- |
0.16 |
0.33 |
|
SET1+SET2 |
0.44 |
0.60 |
0.37 |
5 |
0.64 |
- |
- |
0.51 |
0.35 |
Note. R2 L – R2 of model for learning; MSEL – mean square error of model for learning; Q2 - Q2 of model; N – amount of arguments; MSELOO – mean square error of model for “leave-one-out” method; R2T – R2 of prediction for SET1 or SET2, respectively; SET – standart error of prediction for SET1 or SET2, respectively; R2T (set 3) – R2 of prediction for SET3 (RBI); SET (set 3) – standart error of prediction for SET3 (RBI).