Cross validation crossover verification

Cross-validation is a method of detecting whether a model is overfit. The most commonly used cross validation is the K-fold cross validation.

The specific methods are:

1. Divide the data evenly into k parts, 0,1,2,,,k-1

2. Use 1~k-1 data to train the model, and then use the No. 0 data for verification.

3. The 1th data is then used as the validation data. Make a K loop. I finished k-fold cross validation.

This cross-validation approach is characterized by the fact that all data are validated and trained, without wasting data.

Cross validation crossover verification