我正在实现一个神经网络的分类目的,现在我遇到了交叉验证的问题,我的问题如下。
Do we need to train the model first and then cross validate it (K fold) , or we first cross validate the model ,then pick that model which performs well on unseen data and then train it , Could anybody guide me that how this whole procedure runs , it is getting confused because if training'sdone before or after then what is the role of that training part which is in K Fold ? 先谢谢你。
答案是肯定的.使用K折交叉验证测试模型后,建议使用整个数据来训练模型。