Data from #TrainingSet.txt
Representation of molecular structure by 
* SMILES

Correlation coefficients for the TRAINING SET
 Threshold:Probe  1: Average  :Dispersion
         1:  0.7043:    0.7043:    0.0000

The preferable number of epochs of the Monte Carlo optimization for the TRAINING SET
 Threshold:Probe  1: Average  :Dispersion
         1:      13:     13.00:      0.00

Correlation coefficients for the Calibration SET
 Threshold:Probe  1: Average  :Dispersion
         1:  0.5985:    0.5985:    0.0000

The preferable number of epochs of the Monte Carlo optimization for the Calibration SET
 Threshold:Probe  1: Average  :Dispersion
         1:       6:      6.00:      0.00


Preferable threshold T*=1  Preferable average number of epochs   N*=  6.0   Average R2 for calibration set =   0.5985

Root-mean-square error for the TRAINING SET
 Threshold:Probe  1: Average  :Dispersion
         1:  0.9486:    0.9486:    0.0000

The preferable number of epochs of the Monte Carlo optimization for the TRAINING SET
 Threshold:Probe  1: Average  :Dispersion
         1:      13:     13.00:      0.00

Root-mean-square error for the calibration SET
 Threshold:Probe  1: Average  :Dispersion
         1:  0.9410:    0.9410:    0.0000

The preferable number of epochs of the Monte Carlo optimization for the calibration SET
 Threshold:Probe  1: Average  :Dispersion
         1:       6:      6.00:      0.00

Preferable threshold T*=1  Preferable average number of epochs   N*=  6.0   Average RMSE for calibration set =   0.9410

IMPORTANT:
Ideal result if the preferable threshold is the same for correlation
coefficient of the calibration set and for root-mean-square error of the calibration set.
If they are different, most probably, the threshold preferable according
to correlation coefficient of the calibration set is the indicator of the model
with higher predictive potential.
