confusion

confusion(act,pred,pos) returns the confusion matrix determined by the column vectors of same length act and pred. Pos determines which variable is considered to be the positive one.
  • [conf mp] = confusion(act,pred,pos) with conf as a confusion matrix and mp as a vector containing the accurancy, precision, F1score, TPR and FPR of the model performance.

Example

>> act = {'T';'F';'T';'T';'T';'F'}      
% actual classes

pred = {'T';'F';'T';'T';'T';'T'}        
% predicted classes

[conf mp] = confusion(act,pred,'T')

[act:6x1 cell]
[pred:6x1 cell]
[conf:2x2 double]
[mp:1x2 double]

>> conf#

            4            0
            1            1


>> mp#

     0.833333          0.8

See also

plotconfusion