The Improved Least Square Support Vector Machine Based on Wolf Pack Algorithm and Data Inconsistency Rate for Cost Prediction of Substation Projects

Table 2

The forecasting values and actual values of the test set (unit: yuan/kV·A).

Serial number	Actual value	BPNN	SVM	LSSVM	ILSSVM	WPA-DIR-ILSSVM
67	271.7	286.4	287.1	284.6	280.5	268.6
68	262.6	251.3	273.1	255.3	269.6	264.2
69	253.5	270.0	239.6	264.0	259.6	255.9
70	244.5	225.6	254.5	259.5	239.7	247.2
71	235.4	251.4	245.9	246.5	243.1	236.7
72	326.3	335.6	335.2	341.1	335.7	330.5
73	217.2	208.3	201.8	203.6	211.6	215.6
74	208.1	219.0	216.5	218.0	213.9	209.5
75	199.1	210.2	195.2	208.6	204.9	200.6
76	390.0	376.9	395.6	398.5	381.6	387.6
77	280.9	296.1	294.8	269.6	285.6	280.5
78	285.1	302.8	269.6	296.9	295.0	288.0
79	289.4	271.9	302.1	303.2	280.2	293.1
80	293.7	310.5	281.0	285.6	302.8	291.3
81	297.9	313.7	312.3	311.1	307.6	298.1
82	402.2	412.5	389.5	385.6	413.6	405.9
83	306.4	325.3	321.6	322.6	303.7	309.7
84	310.7	298.7	327.0	292.7	320.2	309.1
85	274.9	290.3	287.2	288.6	283.2	276.7
86	319.2	338.1	302.9	323.5	307.5	318.6
87	283.4	300.1	295.7	270.6	293.2	283.6
88	369.5	350.2	385.9	387.1	380.9	372.0