CodePudding user response:
(1) the 1-0 loss function (0-1 lossfunction) :0-1 loss function is the most simple a loss function, more applicable to the classification problem of
(2) the sense of Loss function (Perceptron Loss) :
Sense of loss function is the improvement of 0-1 loss function, it will not like a 1-0 loss function strictly
(3) square loss function (quadratic loss function) :
Square loss function refers to the square of the predicted values and the real value difference, the greater the loss, shows that the predicted values and the real value of the difference, the greater the squared loss function for linear regression tasks more
(4) Hinge loss function (Hinge loss function) :
Hinge loss function is usually applied to binary classification scenario, which can be used to solve the problem of interval maximization