Why in SVM_Accuracy function Sim_Trn is calculated as: `Sim_Trn = trainset * M * trainset';` and not `Sim_Trn = (trainset * M) * (trainset * M)';` Shouldn't the training kernel be between transformed source inputs?
Why in SVM_Accuracy function Sim_Trn is calculated as:
Sim_Trn = trainset * M * trainset';and not
Sim_Trn = (trainset * M) * (trainset * M)';Shouldn't the training kernel be between transformed source inputs?