Forecasting the length-of-stay of pediatric patients in hospitals: a scoping review

Table 6 Cross-validation modeling approaches and performances

Cross-validation modeling approaches and performances	Reference
Cross-validation modeling approaches and performances	[32]	[42]	[29]	[27]	[38]	[35]	[21]	[36]	[22]	[34]	[24]	[41]	[26]	[3]	[31]	[2]	[30]	[43]	[40]	[33]	[20]	[25]	[23]	[28]	[37]	[19]	[18]	[39]	Frequency
Traditional Holdout		✔					✔	✔	✔					✔	✔	✔												✔	8
Temporal Holdout				✔									✔			✔							✔			✔			5
K-fold				✔																							✔		2
Performance (R²)	0.43	0.9415	0.63–0.82	0.66–0.79	0.36–0.43		0.38	0.097–0.237		0.242–0.278	0.66			0.22–0.30		0.694–0.831		0.329	0.04–0.225	0.47	0.4464	0.78	0.7027	0.51	0.1287				19
Performace (RMSE)													6.2–18.8			0,296 - 0,588													2
Performace (MAE)													4.2–14.6													2.5–4.26			2
Performance (Variance)						30.7–57%																							1
Performace (Average MSE)									0.08–0.38																				1
Performance (% Prediction up to 12 h)															27–46%														1
Performance (Correlation between forecasting models)																	0.92												1
Performance (% *Perfect Prediction)																										12.9–21.2%			1
Performance (% *Predictons up to one day)																										34–51.4%			1
Performance (Mean correct performance)																											60–80%		1
Performance (*Correlation between the predicted and actual LSPPH)																												0.85–0.92	1

ISSN: 1472-6963