Prediksi Konsumsi Rokok di Indonesia dengan Single Exponential Smooting

Abstract

This study predicts cigarette consumption per capita in Medan City using the Single Exponential Smoothing (SES) model. The problem faced is the fluctuation and increasing trend of cigarette consumption which requires accurate prediction. The SES model was chosen due to its ability to capture historical data and produce accurate forecasts. To test the accuracy of the model, three prediction error indicators were used, namely Mean Squared Error (MSE), Mean Absolute Deviation (MAD), and Mean Absolute Percentage Error (MAPE). The results of the analysis show that the SES model has low MSE, MAD, and MAPE values, which indicates a high level of accuracy. For the type of filter clove cigarettes, the MAPE value is 9.43%, for clove cigarettes without filter, the MAPE value is 27.77%, white cigarettes MAPE value is 27.73%, tobacco MAPE value is 38.13% and for other types of cigarettes and tobacco 33.20% where the accuracy results are good.

The predictions generated by this model show an increasing trend of consumption in Medan City. These predictions provide important insights for policymakers in designing cigarette consumption control. This study thus confirms the effectiveness of the SES model in predicting per capita cigarette consumption in Indonesia. It makes a significant contribution to the literature on tobacco product consumption prediction.