resubPredict

Class: ClassificationDiscriminant

Predict resubstitution labels of discriminant analysis classification model

Syntax

label = resubPredict(obj)
[label,posterior] = resubPredict(obj)
[label,posterior,cost] = resubPredict(obj)


Description

label = resubPredict(obj) returns the labels obj predicts for the data obj.X. label is the predictions of obj on the data that fitcdiscr used to create obj.

[label,posterior] = resubPredict(obj) returns the posterior class probabilities for the predictions.

[label,posterior,cost] = resubPredict(obj) returns the predicted misclassification costs per class for the resubstituted data.

Input Arguments

obj	Discriminant analysis classifier, produced using fitcdiscr.

Output Arguments

label	Response obj predicts for the training data. label is the same data type as the training response data obj.Y. The predicted class labels are those with minimal expected misclassification cost; see Prediction Using Discriminant Analysis Models.
posterior	N-by-K matrix of posterior probabilities for classes obj predicts, where N is the number of observations and K is the number of classes.
cost	N-by-K matrix of predicted misclassification costs. Each cost is the average misclassification cost with respect to the posterior probability.

Examples

Find the total number of misclassifications of the Fisher iris data for a discriminant analysis classifier:

load fisheriris
obj = fitcdiscr(meas,species);
Ypredict = resubPredict(obj); % the predictions
Ysame = strcmp(Ypredict,species); % true when ==
sum(~Ysame) % how many are different?

ans =
     3

More About

Posterior Probability

posterior(i,k) is the posterior probability of class k for observation i. For the mathematical definition, see Posterior Probability.