tensorflow. I understand the concept of _ZC

1, convolutional neural network (CNN) and recurrent neural network (RNN)

2, loss of function in the role of counter-propagating - Simple Book .html ( https://www.jianshu.com/p/f987103ec046 )

 As can be seen from the loss function are basically the real and predicted values of two parts, the proper loss function, you can play to make predictive value has been close to the true value of the effect, when the predicted value and the actual value is equal to, loss value of the minimum .

  ZC: I should be able to be understood: the smaller the loss of the value returned by training, should be that the better the training

3、

4、

5、

 

Guess you like

Origin www.cnblogs.com/pythonzc/p/12234139.html