Table 6.

Fitting value of each variable after removing abnormal points from test data.

Seq	Residual ($\epsilon (j)$)	$y(j)$	$\widehat{y}(j)$	$f_1(j)$	$f_2(j)$	$f_3(j)$	$f_4(j)$
1	13.1543	13	− 0.1543	0.469	0.6358	0	0.7883
2	− 1.9374	0	1.9374	0.469	0.5012	0	0.926
3	1.8919	0	− 1.8919	0.814	0	0.434	0.2755
4	–	–	–	–	–	–	–
5	− 2.3402	7	9.3402	1.4768	0.5012	0.4797	1
6	–	–	–	–	–	–	–
7	7.6711	0	− 7.6711	0.7384	0.6653	0.5154	0.2755
8	–	–	–	–	–	–	–
9	− 0.0057	5	5.0057	1.5202	0.6358	0.434	0.5128
10	− 4.9586	0	4.9586	0.7384	0.6523	0	1
11	− 5.4694	2	7.4694	0.7384	0.6523	0.0277	1.3011
12	14.6042	21	6.3958	1.4768	0.7704	0.9137	1.3751
13	− 3.2508	19	22.2508	2.2152	0.5502	0.5818	1.5868
14	− 9.9350	0	9.935	1.6195	0.5012	0.434	0.8264
15	− 3.1375	3	6.1375	1.2074	0.5012	0.7296	1.3011
16	− 5.7666	0	5.7666	1.6764	1.2716	0.434	0.7985
17	–	–	–	–	–	–	–
18	3.5092	0	− 3.5092	0.7384	0.5012	0	0
19	− 1.6303	0	1.6303	0.814	0.6523	0.1478	0.7245
20	6.6882	18	11.3118	1.4768	0.6523	0.0457	0.7883
21	− 5.0519	1	6.0519	2.2152	1.137	1.3948	1.2015
22	− 6.1409	1	7.1409	2.2152	1.137	0.868	0.6887
23	− 4.1729	3	7.1729	1.3501	1.6677	0	1.0996
24	6.7371	0	− 6.7371	0.469	0.5012	0	0
25	− 3.7032	14	17.7032	1.3501	0	0	1.1633
26	− 3.1131	0	3.1131	0.7384	1.0024	0	1.0256
27	− 1.6942	2	3.6942	1.4768	1.7332	0.9266	1.7143
28	− 1.0520	0	1.052	0.469	0	0	0.5128
29	0.7189	13	12.2811	1.6195	0.1511	0.6095	1.0638
30	8.1202	33	24.8798	3.4226	2.8407	1.6468	3.0512
31	− 4.7308	8	12.7308	1.4768	0.3347	0.434	1.2015
32	8.0536	39	30.9464	2.5575	0.049	0	1.0638
33	3.2505	72	68.7495	4.8994	1.8058	0.6974	4.0536
34	− 6.3088	20	26.3088	2.8838	1.6872	0.9494	2.3267