Validation set vs Test set
as you said that :
- the purpose of validation is to make sure that parameters don't overfit
- purpose of test is to make sure that hyperparameters don't overfit
so does it mean that in a classification problem (out of neural network) [example for a not really complexe binary classification that doesn't have parameters and optimizer (eg. xgboost model)] we don't need to split our data to train, validation and test ? and that splitting it to train and test is sufficient in this case ?
- the purpose of validation is to make sure that parameters don't overfit
- purpose of test is to make sure that hyperparameters don't overfit
so does it mean that in a classification problem (out of neural network) [example for a not really complexe binary classification that doesn't have parameters and optimizer (eg. xgboost model)] we don't need to split our data to train, validation and test ? and that splitting it to train and test is sufficient in this case ?
0 answers ( 0 marked as helpful)
Submit an answer
