This file contains average values for statistical characteristics which were obtained
in 1 probes of the Monte Carlo optimization
SMILES taken from file #TrainingSet.txt
Method taken from file Method.txt

SMILES is used for the model
Threshold from 1 to 1

n is the number of compounds in set;
r is Correlation coefficient;
s is standard error of estimation;
F is Fischer F-ratio.
ns, r2s, ss, and Fs are statistical characteristics of Training set.
nc, r2c, sc, and Fc are statistical characteristics of invisible-Training set.
nt, r2t, st, and Ft are statistical characteristics of calibration set.
Rm2av is the average of Rm2 metric, it should be > 0.5 [1]
[1] PK Ojha,I Mitra, RN Das,K Roy,Chemometr Intell Lab 107(2011)194-205
Number of epochs of optimization is 25
Number of probes of optimization is 1
Weight for dr in balance of correlations is 0,1
Start step of the optimization is 0,5*CW(SA)
Precision of the optimization is 0,1*CW(SA)
CW(SA) is weight of SA at the start

N111 is the number of SA which take place in training, 
in invisible training, and in calibration sets
W% is ratio of N111/(number of all SA)
N110 is number of SA which take place in Training and invisible-Training
N101 is number of SA which take place in Training and calibration
N100 is number of SA which take place in Training (only)

Model 
Involves from SMILES:
S(k)
SS(k)
HARD
Balance of correlations i.e. Training-Invisible training-Calibration system
The dr-weight is 0,1


Trshd:Nact :  ns :    rs2:       ss :    fs :  nc :    rc2:       sc :    fc :  nt :    rt2:       st :    ft : Rm2av : W% :N111:N110:N101:N100:SplitDefect
    1:  110:   39:    0.7214:     0.487:     96:   27:    0.8702:     0.769:    168:   17:    0.8250:     0.492:     71:    0.7408:  71:  78:  15:   6:  11:    26.43120
