Table 3.

Performance of various autoMLs and traditional ML algorithms for predicting mRS at 3 months.

Sr. No.	Models	Training accuracy	Testing accuracy	AUC	Sensitivity	Specificity
1	Auto-Gluon	93.12	83.95	0.96	0.75	0.85
2	MLJAR logloss	0.15*	85.18	0.91	0.83	0.91
3	Auto-Sklearn	96.6	82.7	0.89	0.75	0.88
4	TPOT	96	76.54	0.9	0.7	0.85
5	H2O	70.67	72.8	NA*	0.63	0.83
6	Decision tree classifier	99.99	69.13	0.76	0.72	0.85
7	Logistic regression	91.33	72.83	0.89	0.69	0.84
8	Random forest	98.45	72.83	0.89	0.68	0.84
9	kNN	92	51.85	0.67	0.42	0.72
10	SVM	93.11	76.54	0.87	0.72	0.86

The values of AUC, sensitivity, and specificity are for the testing set. ^*Due to the technical limitation of the concerned autoML, the values could not be calculated.