Validation in prediction research: the waste by data splitting

In conclusion, random data splitting should be abolished for validation of prediction models.

Source: Journal of Clinical Epidemiology - July 28, 2018 Category: Epidemiology Authors: Ewout W. Steyerberg Tags: Commentary Source Type: research