Page 449 - Contributed Paper Session (CPS) - Volume 4

CPS2564 Tiffany Rizkika et al.

2) Statistical Filtering-based Model
Here is MAPE value of IQR.Spline dan KDE.Spline Model :
MAPE
MAPE
Commodity IQR Smooth KDE Smooth Spline
Spline
(1) (2) (3)
Long Bean 16,95 17,83
Mackerel 18,01 25,04

Instant Dry 25,62 0,81
Noodles
Peanuts 118,42 62,33
Vegetable Tomato 15,33 15,03
Based on table above, IQR.Spline Model is better for commodities Long
Bean and Mackerel, while KDE.Spline is better for commodities Instant Dry
Noodles, Peanuts, and Vegetable Tomato.
3) Time series-based Model
Here is MAPE value of time series-based model:

MAPE Value
Komoditas Model Model Model Model Model Model
3 4 5 6 7 8
(1) (2) (3) (4) (5) (6) (7)
Long Bean 16,06 16,06 25,73 18,95 16,40 27,29
Mackerel 34,69 34,69 34,69 23,05 22,14 21,75
Instant Dry
Noodles 20,43 20,43 27,60 1,18 0,97 10,34
Peanuts 106,46 106,46 145,41 62,01 49,03 71,92
Vegetable 25,93 25,93 26,78 26,19 13,02 27,04
Tomato

438 | I S I W S C 2 0 1 9