| dc.description.abstract | Exponential smoothing basically pave the past data by performing weighting decreases
exponentially with observed values of older or value of newer assigned a weighting that is
relatively larger than the value of observation is longer and consider the random influences,
trends, seasonality on past data that will be smoothed in the selection method used yaang. By
analyzing the pattern of data on sales data PT Central protein Prima, Tbk using Plot Data,
Correlation Coefficient, and Test Seasonal then the pattern data contained are data patterns
seasonal so used the smoothing method of exponentially three parameters: Methods Winter in
the completion of the problem in this study. To suppress the forecasting error made trial and
error to obtain the smallest MSE value of value parameters are used. The smallest MSE value
obtained for 66. 131.08 with parameter values α = 0.4; β = 0.1; and ɣ = 0.1. By using the
exponential smoothing three parameters: Methods Winter of the obtained equation to predict
the next season as follows.
m m F xm I 60 48 1.263,66 1,32
By entering a value m = the period ahead, the obtained value of forecasting: January =
1168.58, February = 1346.10, March =1214.16, April = 1253.96, May = 1475.51, June =
1496.75, July = 1420.12, August = 1.365, 69, September = 1247.89, October =1305.28,
November= 1303.06 , and December = 1,091, 99. Comparison of the accuracy of forecasting
using exponential smoothing three parameters: Methods Winter with Subjective Methods is
used for this is to compare the value of MAPE obtained from each forecasting. By using the
exponential smoothing three parameters: Methods Winter obtained MAPE by 15, 25, whereas
with forecasting subjective method is used for this value amounted to 28.93 MAPE so that it
can be concluded exponential smoothing three parameters: Methods Winter has a better
forecasting accuracy. | en_US |
