Notes - Evaluation Model: K-fold cross-validation

A high data availability assessment model training method

  • Training assessment model
  • Super model parameter choices

Implementation process is not unique, mainly on the utilization of data

现在对模型进行训练,评估
采用10折交叉验证方法,需要对模型训练10次,测试10次,计算10次相关指标
将数据集切分成10份数据集,
第一次保留第一份数据集为测试集,剩下的数据用来训练模型
第二次保留第二份数据集为测试集,剩下的数据用来训练模型
...

data:

Machine learning model selection (K-fold cross validation, the hyper-parameters selected
K-fold cross validation principle and achieve python

Guess you like

Origin blog.csdn.net/chen_holy/article/details/91564786