Indonesia has a discrepancy between the realized amount and the planned amount of gasoline consumption. This discrepancy has burdened the national budget in Indonesia because this unplanned increase in the imported amount of gasoline has made the government pay more to the importers. The objectives of this research are to develop a robust and accurate model to forecast future gasoline consumption and to provide an attractive alternative model for gasoline consumption forecasting by applying a metaheuristic approach. We apply the harmony search (HS) algorithm for developing a model of gasoline consumption in Indonesia using general socioeconomic variables that can be easily retrieved from public data. The variables used are the gross domestic product (GDP), population, and the total numbers of passenger cars and motorcycles. The HS algolithm selects the optimal weight factors within the proposed exponential model. The results show that the proposed exponential HS algorithm-based model outperforms the conventional nonlinear regression method and particle swarm optimization (PSO)-based model in terms of the mean absolute percentage error (MAPE).
ASJC Scopus subject areas
- Artificial Intelligence
- Human-Computer Interaction
- Computer Vision and Pattern Recognition