past.Evaluation.svm_cross_validation

past.Evaluation.svm_cross_validation(mtx, target, Kfold=5)

K-fold cross validation, taking low-dimensional embedding as input, annotation as output and SVM with rbf kernel as classifier

Parameters:

• mtx – latent feature matrix

• target – annotation

• Kfold – number of fold for cross validation

Returns:

• Acc – Accuracy of cross validation

• K – Kappa of cross validation

• mF1 – mF1 of cross validation

• wF1 – wF1 of cross validation