An Approach for Demand Forecasting in Steel Industries Using Ensemble Learning
Table 2
Statistical description of influence features for forecasting the demand.
Indices
Outcome
Mean
8231.97
52444.08
167.42
28.93
1.46
60.09
51205.62
S.D.
2269.08
12385.54
43.10
1.28
0.90
9.68
10848.99
6985.25
43731.75
135.00
28.00
1.00
52.00
45167.18
7883.50
52941.50
150.00
29.00
1.00
62.00
52418.97
9640.50
59861.00
200.00
30.00
2.00
67.25
61496.95
Range
(4010, 10178)
(30275, 60000)
(100, 200)
(26, 30)
(0, 4)
(38, 76)
(26072, 62275)
Skewness
0.297
0.123
0.454
0.322
1.403
−0.309
−0.624
Kurtosis
−0.520
−0.801
−0.740
2.569
1.180
−0.829
−0.553
Data type
Numeric
Numeric
Numeric
Numeric
Numeric
Numeric
Numeric
S.D. = standard deviation, = availability, = raw material, = worker, = working day, = holiday, = down time, outcome = demand level, = first quantile, = second quantile, = third quantile.
