BioinformaticsbioRxivPreprint
Exploring an experiment-split method to estimate the generalization ability in new data: DeepKme as an example
The paper introduces the experiment-split method to address overfitting in cross-validation by splitting data based on experimental sources, using DeepKme as an example to demonstrate its effectiveness.