Random
In this type of split, the input data in randomly assigned to the test, train and validation set according to the given sizes.
The split is done without replacement, which means that the validation, test and training set doesn't contain the same data.
Random split is the simplest split strategy and is effective in most of the cases.
Get AI confident. Start using Hasty today.
Our platform is completely free to try. Sign up today to start your two-month trial.
On the 9th of February, we are hosting a ML-IRL event with speakers from Bayer, Intel, and Infineon.