Machine Learning Approaches to Predict Patient’s Length of Stay in Emergency Department

Table 7

Comparison between the 3-CAT and 5-CAT for all algorithms.

Classifier	3-CAT				5-CAT
	Correctly classified		Incorrectly classified		Correctly classified		Incorrectly classified
Logistic regression	322	80.5%	78	19.5%	242	60.5%	158	39.5%
Naive Bayes	327	81.8%	73	18.2%	263	65.8%	137	34.2%
REP tree	345	86.3%	55	13.7%	252	63.0%	148	37.0%
SMO	333	83.3%	67	16.7%	254	63.5%	146	36.5%
lazy.IBk	310	77.5%	90	22.5%	211	52.8%	189	47.2%
Decision stump	343	85.8%	57	14.2%	258	64.5%	142	35.5%
Random Forest	330	82.5%	70	17.5%	235	58.8%	165	41.3%
MLP (0.0001)	332	83.0%	68	17.0%	212	53.0%	188	47.0%
MLP (0.001)	332	83.0%	68	17.0%	254	63.5%	146	36.5%
MLP (0.01)	317	79.3%	83	20.7%	240	60.0%	160	40.0%