Mostly, we divide the whole set into 3 parts.

1. Training Set

Train your model. You can optimize your model, change some parameters and so on.

2. Validation Set

Look at your models and select the best performing approach using the validation data. Compare your algorithms and their training parameters and decide on a winner.

3. Test Set

Do not change any parameters, just test your winner model

Reference

  1. http://stats.stackexchange.com/questions/19048/what-is-the-difference-between-test-set-and-validation-set
  2. http://stats.stackexchange.com/questions/9357/why-only-three-partitions-training-validation-test
Logo

CSDN联合极客时间,共同打造面向开发者的精品内容学习社区,助力成长!

更多推荐