1
vote
splitting small dataset into train and test set/
This is a common problem:
Evaluation on test data (i.e. data not seen during the training) is crucial, especially with few samples, the rist of overfitting is significant and unseen data helps to ...
- 1,717
Only top scored, non community-wiki answers of a minimum length are eligible