Very Short-Term Load Forecasting Using Hybrid Algebraic Prediction and Support Vector Regression

Table 2

Training set from the months July, August, and September 2014.

Date
	July 1	2	0.920	−0.375	1341.633
July 3	−7	−2.710	7.625	−4110.892
July 9	4	1.540	5.167	1642.079
July 16	−1	−0.290	−1.625	−818.390
July 17	−2	−0.330	4.667	−1285.362
July 23	−2	−2.040	4.833	−2228.795
July 24	−1	−1.960	−0.250	−915.093
July 25	9	5.540	−5.500	3170.716
July 28	−4	−2.080	−3.958	−4190.624
Aug 4	−6	−3.790	25.208	−7426.996
Aug 12	2	1.917	0.458	2501.683
Aug 14	−3	−1.670	−3.875	−3083.127
Aug 27	−2	−1.208	−2.542	−1438.378
Sept 1	2	0.580	1.625	2719.843
Sept 2	−2	−0.460	−3.625	−1321.296
Sept 12	−1	−0.500	6.208	−1582.431
Sept 17	−4	−0.708	7.417	−2069.028