Table 3.

Combinations of filters and metrics with a cost (w_L=3, w_E=1) of -0.5 or less along with values for no filtering and the Stampe and Savitzky-Golay filters.

	Filter	Win (ms)	Dispersion	fLat (ms)	Error (deg)	Precision	Cost
			Metric	Thr	Rem			RMS	SD	Extent
1	Kalman	500	STD	0.5°	95%	0	1.21	0.13	0.16	0.21	-0.565
2	Kalman	500	M-M	0.05°	95%	0	1.20	0.14	0.19	0.25	-0.546
3	Tri	500	STD	1.0°	95%	2.0	1.14	0.10	0.15	0.18	-0.545
4	Tri	500	M-M	0.05°	95%	0.3	1.14	0.11	0.16	0.19	-0.545
5	Tri	500	Rad	0.05°	95%	2.0	1.16	0.10	0.15	0.18	-0.542
6	Kalman	500	STD	1.0°	95%	3.2	1.13	0.10	0.14	0.17	-0.540
7	Kumar	500	STD	0.5°	95%	2.1	1.20	0.12	0.15	0.19	-0.535
8	Kumar	500	Rad	0.05°	95%	3.9	1.12	0.11	0.15	0.18	-0.531
9	Kumar	500	M-M	0.05°	95%	5.8	1.14	0.10	0.14	0.17	-0.525
10	Tri	500	STD	0.5°	95%	3.1	1.21	0.13	0.16	0.21	-0.522
11	Kalman	500	Rad	0.05°	95%	2.1	1.20	0.13	0.19	0.23	-0.517
12	Kumar	500	STD	0.5°	50%	7.6	1.23	0.07	0.16	0.18	-0.514
13	Average	500	STD	1.0°	95%	5.0	1.17	0.10	0.18	0.21	-0.514
14	Tri	500	STD	0.5	50%	6.5	1.25	0.09	0.17	0.19	-0.512
15	Kumar	500	STD	1.0	95%	10.9	1.17	0.08	0.13	0.15	-0.509
16	Kalman	500	Rad	0.1	95%	8.2	1.13	0.10	0.16	0.19	-0.505
…
125	Stampe					11.2	1.19°	0.26°	0.33°	0.42°	-0.372
…
275	Savitzky-Golay			13.6	1.22°	0.40°	0.40°	0.58°	-0.281
…
334	No filter					0	1.25°	0.63°	0.48°	0.80°	-0.245