Skip to main content
. 2023 Apr 3;61(9):2227–2240. doi: 10.1007/s11517-023-02802-5

Fig. 4.

Fig. 4

Validation scheme, where all the instances of one single patient are included in test set. The remaining instances are used to perform 5-fold CV and hyperparameter tuning. The split is random, 80% and 20 % to train and validation, respectively. In total, six models are tested