multidim

This is analysis that compares several groups.

Computations cover several cases.

min

This reduction is the minimum.
Computations cover several cases.
acc
This measure is the minimum of the accuracy.
✔ 0.959 min acc
Obtained from 12 values
pr
This measure is the minimum of the positive rate.
? 0.000 min pr
Obtained from 12 values
tpr
This measure is the minimum of the true positive rate/recall/sensitivity/hit rate.
X 0.000 min tpr
Obtained from 12 values
tnr
This measure is the minimum of the true negative rate/specificity.
X 0.000 min tnr
Obtained from 12 values
ppv
This measure is the minimum of the positive predictive value/precision.
X 0.000 min ppv
Obtained from 12 values
f1
This measure is the minimum of the f1 score.
X 0.000 min f1
Obtained from 12 values
gmi
This measure is the minimum of the geometric mean of tpr and tnr - accounts for class imbalance.
X 0.000 min gmi
Obtained from 12 values
tar
This measure is the minimum of the true acceptance ratio (true positives compared to all).
? 0.000 min tar
Obtained from 12 values
trr
This measure is the minimum of the true rejection ratio (true negatives compared to all).
? 0.000 min trr
Obtained from 12 values
lift
This measure is the minimum of the lift ratio (tpr divided by pr).
? 0.000 min lift
Obtained from 12 values
mcc
This measure is the minimum of the Matthews correlation coefficient.
X 0.000 min mcc
Obtained from 12 values
kappa
This measure is the minimum of the Cohen's Kappa score.
X 0.000 min kappa
Obtained from 12 values

max

This reduction is the maximum.
Computations cover several cases.
pr
This measure is the maximum of the positive rate.
? 1.000 max pr
Obtained from 12 values
tar
This measure is the maximum of the true acceptance ratio (true positives compared to all).
? 1.000 max tar
Obtained from 12 values
trr
This measure is the maximum of the true rejection ratio (true negatives compared to all).
? 1.000 max trr
Obtained from 12 values
lift
This measure is the maximum of the lift ratio (tpr divided by pr).
? 11.000 max lift
Obtained from 12 values

maxerror

This reduction is the maximum deviation from the ideal value.
Computations cover several cases.
acc
This measure is the maximum deviation from the ideal value of the accuracy.
✔ 0.041 maxerror acc
Obtained from 12 values
tpr
This measure is the maximum deviation from the ideal value of the true positive rate/recall/sensitivity/hit rate.
X 1.000 maxerror tpr
Obtained from 12 values
tnr
This measure is the maximum deviation from the ideal value of the true negative rate/specificity.
X 1.000 maxerror tnr
Obtained from 12 values
ppv
This measure is the maximum deviation from the ideal value of the positive predictive value/precision.
X 1.000 maxerror ppv
Obtained from 12 values
f1
This measure is the maximum deviation from the ideal value of the f1 score.
X 1.000 maxerror f1
Obtained from 12 values
gmi
This measure is the maximum deviation from the ideal value of the geometric mean of tpr and tnr - accounts for class imbalance.
X 1.000 maxerror gmi
Obtained from 12 values
mcc
This measure is the maximum deviation from the ideal value of the Matthews correlation coefficient.
X 1.000 maxerror mcc
Obtained from 12 values
kappa
This measure is the maximum deviation from the ideal value of the Cohen's Kappa score.
X 1.000 maxerror kappa
Obtained from 12 values

wmean

This reduction is the weighted average.
Computations cover several cases.
acc
This measure is the weighted average of the accuracy.
✔ 0.967 wmean acc
Obtained from 12 values
pr
This measure is the weighted average of the positive rate.
? 0.485 wmean pr
Obtained from 12 values
tpr
This measure is the weighted average of the true positive rate/recall/sensitivity/hit rate.
✔ 1.000 wmean tpr
Obtained from 12 values
tnr
This measure is the weighted average of the true negative rate/specificity.
✔ 0.936 wmean tnr
Obtained from 12 values
ppv
This measure is the weighted average of the positive predictive value/precision.
✔ 0.932 wmean ppv
Obtained from 12 values
tar
This measure is the weighted average of the true acceptance ratio (true positives compared to all).
? 0.453 wmean tar
Obtained from 12 values
trr
This measure is the weighted average of the true rejection ratio (true negatives compared to all).
? 0.515 wmean trr
Obtained from 12 values
lift
This measure is the weighted average of the lift ratio (tpr divided by pr).
? 2.176 wmean lift
Obtained from 12 values
kappa
This measure is the weighted average of the Cohen's Kappa score.
✔ 0.931 wmean kappa
Obtained from 12 values

mean

This reduction is the average.
Computations cover several cases.
acc
This measure is the average of the accuracy.
✔ 0.981 mean acc
Obtained from 12 values
pr
This measure is the average of the positive rate.
? 0.408 mean pr
Obtained from 12 values
tpr
This measure is the average of the true positive rate/recall/sensitivity/hit rate.
✔ 0.917 mean tpr
Obtained from 12 values
tnr
This measure is the average of the true negative rate/specificity.
✔ 0.886 mean tnr
Obtained from 12 values
ppv
This measure is the average of the positive predictive value/precision.
✔ 0.866 mean ppv
Obtained from 12 values
f1
This measure is the average of the f1 score.
✔ 0.890 mean f1
Obtained from 12 values
gmi
This measure is the average of the geometric mean of tpr and tnr - accounts for class imbalance.
✔ 0.891 mean gmi
Obtained from 12 values
tar
This measure is the average of the true acceptance ratio (true positives compared to all).
? 0.389 mean tar
Obtained from 12 values
trr
This measure is the average of the true rejection ratio (true negatives compared to all).
? 0.592 mean trr
Obtained from 12 values
lift
This measure is the average of the lift ratio (tpr divided by pr).
? 3.013 mean lift
Obtained from 12 values
mcc
This measure is the average of the Matthews correlation coefficient.
✔ 0.792 mean mcc
Obtained from 12 values
kappa
This measure is the average of the Cohen's Kappa score.
✔ 0.791 mean kappa
Obtained from 12 values

gm

This reduction is the geometric mean.
Computations cover several cases.
acc
This measure is the geometric mean of the accuracy.
✔ 0.981 gm acc
Obtained from 12 values
pr
This measure is the geometric mean of the positive rate.
? 0.000 gm pr
Obtained from 12 values
tpr
This measure is the geometric mean of the true positive rate/recall/sensitivity/hit rate.
X 0.000 gm tpr
Obtained from 12 values
tnr
This measure is the geometric mean of the true negative rate/specificity.
X 0.000 gm tnr
Obtained from 12 values
ppv
This measure is the geometric mean of the positive predictive value/precision.
X 0.000 gm ppv
Obtained from 12 values
f1
This measure is the geometric mean of the f1 score.
X 0.000 gm f1
Obtained from 12 values
gmi
This measure is the geometric mean of the geometric mean of tpr and tnr - accounts for class imbalance.
X 0.000 gm gmi
Obtained from 12 values
tar
This measure is the geometric mean of the true acceptance ratio (true positives compared to all).
? 0.000 gm tar
Obtained from 12 values
trr
This measure is the geometric mean of the true rejection ratio (true negatives compared to all).
? 0.000 gm trr
Obtained from 12 values
lift
This measure is the geometric mean of the lift ratio (tpr divided by pr).
? 0.000 gm lift
Obtained from 12 values
mcc
This measure is the geometric mean of the Matthews correlation coefficient.
X 0.000 gm mcc
Obtained from 12 values
kappa
This measure is the geometric mean of the Cohen's Kappa score.
X 0.000 gm kappa
Obtained from 12 values

pnorm

This reduction is the p-norm (default L2).
Computations cover several cases.
acc
This measure is the p-norm (default L2) of the accuracy.
X 3.400 pnorm acc
Obtained from 12 values
pr
This measure is the p-norm (default L2) of the positive rate.
? 1.666 pnorm pr
Obtained from 12 values
tpr
This measure is the p-norm (default L2) of the true positive rate/recall/sensitivity/hit rate.
X 3.317 pnorm tpr
Obtained from 12 values
tnr
This measure is the p-norm (default L2) of the true negative rate/specificity.
X 3.206 pnorm tnr
Obtained from 12 values
ppv
This measure is the p-norm (default L2) of the positive predictive value/precision.
X 3.136 pnorm ppv
Obtained from 12 values
f1
This measure is the p-norm (default L2) of the f1 score.
X 3.221 pnorm f1
Obtained from 12 values
gmi
This measure is the p-norm (default L2) of the geometric mean of tpr and tnr - accounts for class imbalance.
X 3.223 pnorm gmi
Obtained from 12 values
tar
This measure is the p-norm (default L2) of the true acceptance ratio (true positives compared to all).
? 1.612 pnorm tar
Obtained from 12 values
trr
This measure is the p-norm (default L2) of the true rejection ratio (true negatives compared to all).
? 2.232 pnorm trr
Obtained from 12 values
lift
This measure is the p-norm (default L2) of the lift ratio (tpr divided by pr).
? 14.097 pnorm lift
Obtained from 12 values
mcc
This measure is the p-norm (default L2) of the Matthews correlation coefficient.
X 3.008 pnorm mcc
Obtained from 12 values
kappa
This measure is the p-norm (default L2) of the Cohen's Kappa score.
X 3.003 pnorm kappa
Obtained from 12 values

maxrel

This reduction is the maximum relative difference.
Computations cover several cases.
acc
This measure is the maximum relative difference of the accuracy.
✔ 0.041 maxrel acc
Obtained from 12 values
pr
This measure is the maximum relative difference of the positive rate.
X 1.000 maxrel pr
Obtained from 12 values
tpr
This measure is the maximum relative difference of the true positive rate/recall/sensitivity/hit rate.
X 1.000 maxrel tpr
Obtained from 12 values
tnr
This measure is the maximum relative difference of the true negative rate/specificity.
X 1.000 maxrel tnr
Obtained from 12 values
ppv
This measure is the maximum relative difference of the positive predictive value/precision.
X 1.000 maxrel ppv
Obtained from 12 values
f1
This measure is the maximum relative difference of the f1 score.
X 1.000 maxrel f1
Obtained from 12 values
gmi
This measure is the maximum relative difference of the geometric mean of tpr and tnr - accounts for class imbalance.
X 1.000 maxrel gmi
Obtained from 12 values
tar
This measure is the maximum relative difference of the true acceptance ratio (true positives compared to all).
X 1.000 maxrel tar
Obtained from 12 values
trr
This measure is the maximum relative difference of the true rejection ratio (true negatives compared to all).
X 1.000 maxrel trr
Obtained from 12 values
lift
This measure is the maximum relative difference of the lift ratio (tpr divided by pr).
X 1.000 maxrel lift
Obtained from 12 values
mcc
This measure is the maximum relative difference of the Matthews correlation coefficient.
X 1.000 maxrel mcc
Obtained from 12 values
kappa
This measure is the maximum relative difference of the Cohen's Kappa score.
X 1.000 maxrel kappa
Obtained from 12 values

maxdiff

This reduction is the maximum difference.
Computations cover several cases.
acc
This measure is the maximum difference of the accuracy.
✔ 0.041 maxdiff acc
Obtained from 12 values
pr
This measure is the maximum difference of the positive rate.
X 1.000 maxdiff pr
Obtained from 12 values
tpr
This measure is the maximum difference of the true positive rate/recall/sensitivity/hit rate.
X 1.000 maxdiff tpr
Obtained from 12 values
tnr
This measure is the maximum difference of the true negative rate/specificity.
X 1.000 maxdiff tnr
Obtained from 12 values
ppv
This measure is the maximum difference of the positive predictive value/precision.
X 1.000 maxdiff ppv
Obtained from 12 values
f1
This measure is the maximum difference of the f1 score.
X 1.000 maxdiff f1
Obtained from 12 values
gmi
This measure is the maximum difference of the geometric mean of tpr and tnr - accounts for class imbalance.
X 1.000 maxdiff gmi
Obtained from 12 values
tar
This measure is the maximum difference of the true acceptance ratio (true positives compared to all).
X 1.000 maxdiff tar
Obtained from 12 values
trr
This measure is the maximum difference of the true rejection ratio (true negatives compared to all).
X 1.000 maxdiff trr
Obtained from 12 values
lift
This measure is the maximum difference of the lift ratio (tpr divided by pr).
X 11.000 maxdiff lift
Obtained from 12 values
mcc
This measure is the maximum difference of the Matthews correlation coefficient.
X 1.000 maxdiff mcc
Obtained from 12 values
kappa
This measure is the maximum difference of the Cohen's Kappa score.
X 1.000 maxdiff kappa
Obtained from 12 values

gini

This reduction is the gini coefficient.
Computations cover several cases.
acc
This measure is the gini coefficient of the accuracy.
✔ 0.007 gini acc
Obtained from 12 values
pr
This measure is the gini coefficient of the positive rate.
? 0.335 gini pr
Obtained from 12 values
tpr
This measure is the gini coefficient of the true positive rate/recall/sensitivity/hit rate.
✔ 0.083 gini tpr
Obtained from 12 values
tnr
This measure is the gini coefficient of the true negative rate/specificity.
✔ 0.096 gini tnr
Obtained from 12 values
ppv
This measure is the gini coefficient of the positive predictive value/precision.
✔ 0.103 gini ppv
Obtained from 12 values
f1
This measure is the gini coefficient of the f1 score.
✔ 0.094 gini f1
Obtained from 12 values
gmi
This measure is the gini coefficient of the geometric mean of tpr and tnr - accounts for class imbalance.
✔ 0.093 gini gmi
Obtained from 12 values
tar
This measure is the gini coefficient of the true acceptance ratio (true positives compared to all).
? 0.348 gini tar
Obtained from 12 values
trr
This measure is the gini coefficient of the true rejection ratio (true negatives compared to all).
✔ 0.231 gini trr
Obtained from 12 values
lift
This measure is the gini coefficient of the lift ratio (tpr divided by pr).
? 0.412 gini lift
Obtained from 12 values
mcc
This measure is the gini coefficient of the Matthews correlation coefficient.
✔ 0.179 gini mcc
Obtained from 12 values
kappa
This measure is the gini coefficient of the Cohen's Kappa score.
✔ 0.180 gini kappa
Obtained from 12 values

stdx2

This reduction is the standard deviation x2.
Computations cover several cases.
acc
This measure is the standard deviation x2 of the accuracy.
✔ 0.028 stdx2 acc
Obtained from 12 values
pr
This measure is the standard deviation x2 of the positive rate.
? 0.509 stdx2 pr
Obtained from 12 values
tpr
This measure is the standard deviation x2 of the true positive rate/recall/sensitivity/hit rate.
? 0.553 stdx2 tpr
Obtained from 12 values
tnr
This measure is the standard deviation x2 of the true negative rate/specificity.
? 0.536 stdx2 tnr
Obtained from 12 values
ppv
This measure is the standard deviation x2 of the positive predictive value/precision.
? 0.528 stdx2 ppv
Obtained from 12 values
f1
This measure is the standard deviation x2 of the f1 score.
? 0.538 stdx2 f1
Obtained from 12 values
gmi
This measure is the standard deviation x2 of the geometric mean of tpr and tnr - accounts for class imbalance.
? 0.539 stdx2 gmi
Obtained from 12 values
tar
This measure is the standard deviation x2 of the true acceptance ratio (true positives compared to all).
? 0.510 stdx2 tar
Obtained from 12 values
trr
This measure is the standard deviation x2 of the true rejection ratio (true negatives compared to all).
? 0.509 stdx2 trr
Obtained from 12 values
lift
This measure is the standard deviation x2 of the lift ratio (tpr divided by pr).
X 5.471 stdx2 lift
Obtained from 12 values
mcc
This measure is the standard deviation x2 of the Matthews correlation coefficient.
? 0.710 stdx2 mcc
Obtained from 12 values
kappa
This measure is the standard deviation x2 of the Cohen's Kappa score.
? 0.709 stdx2 kappa
Obtained from 12 values