Train/dev set error and bias/variance
The dev and test set should:
0
3
Share
Contributors are:
Who are from:
Tags
Data Science
If you have 10,000,000 examples, how would you split the train/dev/test set?
Size of dev/test set size
