Optimisasi Parameter Metode Holt-Winter Dengan Menggunakan Algoritma Genetika


  • Haris Sriwindono Universitas Sanata Dharma
  • Rama Eka Putranto Universitas Sanata Dharma


Forecasting, Holt-Winter, Optimization, Genetic Algorithm, Crossovers, Mutation


TOM is a vegetable supplier to various Yogyakarta supermarkets. During this time, C. V. TOM was unable to meet market demand due to fluctuations in demand. CV. TOM needs time to extend the product in advance from partner farmers, so it is important that C.V. TOM in order to be able to estimate the demand for vegetables to be delivered in the future. These problems should be investigated using forcasting methods to ensure that C.V. TOM fully meets the needs of the market. The forecasting method used here is the multiplicative Holt-Winter method, as the processed data are not stationary or fluctuating product demand data. In the Holt-Winter algorithm, the parameters α, β, ɣ must be determined first. This determination is random and therefore does not necessarily lead to an accurate forecast. In this study, the parameter values α, β, ɣ ​​are searched using a genetic algorithm. Here we use genetic operators, namely single point crossing and reverse mutation. The fitness function used is the mean absolute percentage error (MAPE), the higher the value, the lower the fitness value. The result of this research  α = 0.1119; β = 0.1884 and ɣ = 0.9186 at 5.0571% of MAPE, which means that the predictions are very accurate. This condition can be achieved using the parameters of the genetic algorithm, namely the probability of mutation/crossover = 0.013, the number of chromosomes = 10 and the number of generations is 500


Abdoun. O, Abouchabaka .J., and Tajani. C. (2011). Analyzing the Performance of Mutation Operators to Solve the Travelling Salesman Problem. Wotic.

Fudianto, M. A., Kusbianto, D., Pemenang, M. U. (2020). Sistem Informasi Peramalan Penjualan Obat Ternak Berbasis Web. (Studi Kasus PT Otasindo Prima Satwa Cabang Surabaya). Seminar Informatika Aplikatif Polinema.

Gen, M., & Cheng. (1997). Genetic Algorithms and Engineering Design. Canada: Jhon Wiley & Sons, Inc.

Goldberg. D, (1989). Genetic Algorithm in Search, Optimization, and Machine Learning.

Holland, J. (1975). Adaptation in Natural and Artificial Systems. University of Michigan Press.

Nindian, P. D. (2020). Implementasi Holt-Winters Exponential Smoothing untuk Peramalan Harga Bahan Pangan di Kabupaten Pamekasan. Digital Zone: Jurnal Teknologi Informasi Dan Komunikasi, 11(2), 223–236. https://doi.org/10.31849/digitalzone.v11i2.4797

Oliver. IM, Smith. DJ, and Holland. JRC. (1987). A study of permutation crossover operators on the traveling salesman problem. International Conference on Genetic Algorithms (ICGA’87) Cambridge, MA.

Putra, E. F., Asdi, Y., & Maiyastri, M. (2019). Peramalan dengan Metode Pemulusan Eksponensial Holt-Winter dan Sarima. Jurnal Matematika UNAND, 8(1), 75. https://doi.org/10.25077/jmu.8.1.75-83.2019

Rosalina, E., Sugiarto, S., & Gamal, M. D. . (2016). Metode Peramalan Holt-Winter Untuk Memprediksi Jumlah Pengunjung Perpustakaan Universitas Riau. Repository FMIPA.

Sriwindono, H., Prima Rosa, P.H., Polina,A.M., Nugroho, R.A. (2017). The Model of Elementary School Teachers Placement in Magelang District by Using Genetic Algorithm. International Conference on Computer Science and Artificial Intelligent (CSAI). Jakarta.https://doi.org/10.1145/3168390.3168414

Zainudin, Z., (2013). Algoritma Genetika: Metode Komputasi Evolusioner untuk Menyelesaikan Masalah Optimasi. ANDI.