Peramalan Harga Batubara Acuan Menggunakan Metode PSOSVR dan IPSOSVR
Keywords:
Coal Price Index, Forecasting, Improved-Particle Swarm Optimization, Mean Absolute Percentage Error, Particle Swarm Optimization, Support Vector RegressionAbstract
Coal is a type of fossil fuel that is often used by industrial companies. A prediction model of coalprices index to see future coal prices is needed, so that industrial companies can allocate funds appropriately to maximize profits and minimize production costs. There are several studies that discuss the prediction of reference coal prices using machine learning, one of which is using support vector regression (SVR). However, this method still has shortcomings in determining the correct parameter values. An optimization algorithm is needed to help determine the right parameter value. Therefore, this study aims to forecast reference coal prices using historical data for the period January 2009 to October 2019, using the support vector regression (SVR) method that is optimized with particle swarm optimization (PSO) and improved-particle swarm optimization (IPSO), which is evaluated using the MAPE forecasting results. Based on research that has been done, the prediction of reference coal prices using the
PSOSVR method produces a MAPE value of 3.911% and the IPSOSVR method produces a MAPE value of 3.916%. Whereas the prediction using SVR parameters that is not optimized produces a MAPE value of 13.388%.
ABSTRAK
Batubara adalah salah satu jenis bahan bakar fosil yang sering dimanfaatkan oleh perusahaan industri. Sebuah model prediksi harga batubara acuan untuk melihat harga batubara acuan di masa yang akan datang sangat diperlukan, sehingga perusahaan industri dapat mengalokasikan dana dengan tepat untuk memaksimalkan keuntungan dan meminimumkan biaya produksi. Terdapat beberapa studi yang membahas tentang prediksi harga batubara acuan menggunakan machine learning yang salah satunya yaitu menggunakan support vector regression (SVR). Namun, metode tersebut masih memiliki kekurangan pada penentuan nilai parameter yang tepat. Dalam proses penentuan nilai parameter yang tepat tersebut diperlukan bantuan algoritma optimasi. Oleh karena itu, penelitian ini bertujuan untuk melakukan peramalan harga batubara acuan menggunakan data historis periode bulan Januari 2009 sampai dengan bulan Oktober 2019, dengan menggunakan metode support vector regression (SVR) yang dioptimasi dengan particle swarm optimization (PSO) dan improved-particle swarm optimization (IPSO). Hasil peramalan kemudian dievaluasi menggunakan MAPE. Berdasarkan penelitian yang telah dilakukan, prediksi harga batubara acuan menggunakan metode PSOSVR menghasilkan nilai MAPE sebesar 3,911% dan metode IPSOSVR menghasilkan nilai MAPE sebesar 3,916%. Sedangkan untuk prediksi menggunakan parameter SVR yang tidak dioptimasi menghasilnya nilai MAPE sebesar 13,388%.
References
Chai, Y. (2011). A Coal Mine Gas Concentration Prediction Method based on Particle Swarm Optimization Support Vector Regression. IEEE, 334- 337.
Cholissodin, I., & Riyandani, I. (2016). Swarm Intelligence. Malang: Fakultas Ilmu Komputer, Universitas Brawijaya. Ditjen Minerba Kementrian ESDM. (2019).
Engelbrecht, A. P. (2007). Computational Intelligence: An Introduction 2nded., West Sussex: John Wiley & Sons Ltd.
Hastie, T., Tibshirani, R., & Friedman, J. (2009). The Elements of Statistical Learning Second. New York: Springer-Verlag.
Hseih, H.-I, Lee, T.-P, & Lee, T.-S. (2011). A Hybrid Particle Swarm Optimization and Support Vector Regression Model for Financial Time Series Forecasting. Int. Journal of Business Administration, 2, 48-56.
Kennedy, J., & Eberhart, R. (1995). Particle Swarm Optimization. Neural Networks., Proceedings., IEEE Int. Conf, 4, pp. 1942– 1948.
Khair, U., Fahmi, H., Al Hakim, S., & Rahim, R. (2017, December). Forecasting error calculation with mean absolute deviation and mean
absolute percentage error. In Journal of Physics: Conference Series (Vol. 930, No. 1, p. 012002). IOP Publishing.
Marini, F., & Walczak, B. (2015). Particle swarm optimization (PSO). A tutorial. IEEE Chemometrics and Intelligent La-boratory Systems, 13.
Novitasari, D., Cholissodin, I., & Mahmudy, W. F. (2016). Hybridizing PSO with SA for Optimizing SVR Applied to Software Effort Estimation. TELKOMNIKA, pp. 245-253.
Qasim, S. Q., & Algamal, Z. Y. (2018). Feature Selection Using Particle Swarm Optimization-based Logistic Regression Model. Chemometrics and Intelligent Laboratory System, 182, Pp 41-46.
Saputro, F. D., (2017). Penerapan Support Vector Regression Dengan Optimasi PSO Dalam Memprediksi Harga Gabah. (Doctoral dissertation,
Universitas Komputer Indonesia).
Smola, A. J., & Scholkopf, B. (2004). A Tutorial On Support Vector Regression. The Journal of Statistics and Computing, 14(3) Pp. 199-222.
Vapnik, V. (1999). An Overview of Statistical Learning Theory. IEEE. Transactions on Neural Networks, 10(5), pp.988–999.
Vijayakumar, S. (1999). Sequential Support Vector Classifiers and Regression. In Proceedings of International Conference on Soft Computing (SOCO‘99), 610–619.
Zhang, K., Qiu, B., & Mu, D. (2016). Low-carbon Logistics Distribution Route Planning With Improved Particle Swarm Optimization Algorithm. School of Economics and Management Beijing Jiaotong University Beijing, China.
Zou, Jifeng, Chenlong, L., & Qiao, L. (2015). Fault Prediction Method based on SVR of Improved PSO. China: School of Information Science and
Engineering, Shenyang Ligong University, Shenyang 110159.
