Table B6.

Confusion matrix and performance values for the KNN-based classification system.

Test/Real	Walk	Jump	Immobile	Run	Up	Down	Cycle	Drive	Instances	TP	TN	FP	FN	Accuracy	Recall	Specificity	Precision	F1-Measure
Walk	6726	65	52	143	137	87	121	196	7527	6726	51,490	801	936	97.10%	87.78%	98.47%	89.36%	88.56%
Jump	161	6879	26	265	36	27	62	39	7495	6879	51,768	616	690	97.82%	90.88%	98.82%	91.78%	91.33%
Immobile	65	26	6719	21	95	64	26	382	7398	6719	52,193	679	362	98.26%	94.89%	98.72%	90.82%	92.81%
Run	86	244	36	6854	65	79	84	49	7497	6854	51,834	643	622	97.89%	91.68%	98.77%	91.42%	91.55%
Up	149	98	32	41	6881	375	36	27	7639	6881	51,317	758	997	97.07%	87.34%	98.54%	90.08%	88.69%
Down	187	21	16	31	504	6870	26	41	7696	6870	51,522	826	735	97.40%	90.34%	98.42%	89.27%	89.80%
Cycle	65	51	58	47	124	78	6748	132	7303	6748	52,254	555	396	98.41%	94.46%	98.95%	92.40%	93.42%
Drive	223	185	142	74	36	25	41	6672	7398	6672	51,689	726	866	97.34%	88.51%	98.61%	90.19%	89.34%
Accumulated	7662	7569	7081	7476	7878	7605	7144	7538	59,953	54,349	414,067	5604	5604	97.66%	90.74%	98.66%	90.66%	90.69%