Table 1. Prediction metrics on held out data for the best performing gradient boosted decision trees model.

Antibiotic	TN	FP	FN	TP	S.PRC	S.RCL	R.PRC	R.RCL	S.FSc	R.FSc	ACC
AMP	24	5	5	118	0.83	0.96	0.83	0.96	0.83	0.96	0.93
AMX	80	8	15	115	0.84	0.93	0.91	0.89	0.87	0.91	0.89
AMC	221	34	29	52	0.89	0.60	0.87	0.64	0.87	0.62	0.81
CTZ	309	4	13	50	0.96	0.92	0.99	0.80	0.97	0.85	0.95
CTX	281	5	6	66	0.98	0.93	0.98	0.922	0.98	0.92	0.97
CXM	273	11	24	68	0.92	0.86	0.96	0.74	0.94	0.79	0.91
CET	38	12	17	85	0.70	0.88	0.76	0.83	0.72	0.85	0.81
GEN	316	1	11	48	0.97	0.98	0.99	0.81	0.98	0.89	0.97
TBM	104	3	7	38	0.94	0.92	0.98	0.84	0.95	0.89	0.93
TMP	73	7	10	62	0.88	0.90	0.91	0.86	0.90	0.88	0.89
CIP	281	10	16	69	0.95	0.87	0.97	0.81	0.95	0.84	0.93

TN: true negatives, FN: false negatives, FP: false positives, TP: true positives, PRC: precision, RCL: recall, S: susceptibility, R: resistance, ACC: accuracy for resistance.