Implementation of the Support Vector Regression Algorithm And Particle Swarm Optimization In Sales Forecasting

Mrs Rahmawati (Program Studi Sistem Informasi, Fakultas Ilmu Terapan Dan Sains, Institut Pendidikan Indonesia)
Riyad Sabilul Muminin (Program Studi Bisnis Digital, Fakultas Ekonomi, Universitas Garut)
Rizi Ramdhani (Program Studi Bisnis Digital, Fakultas Ekonomi, Universitas Garut)
Irdam Denni (Institut Pendidikan Indonesia)

RISTEC : Research in Information Systems and Technology
Vol 1, No 1 (2020): Implementation of Manager Performance Decision Support System Using the Simple A

Abstract: Sales forecasting is an important element in planning business strategies that have the possibility of happening in the future. But how to improve the accuracy of the current forecasting process is still a big question mark for companies. This study uses the Support Vector Regression algorithm because it is one of the forecasting techniques that is categorized as good compared to other techniques. The Particle Swarm Optimization algorithm is integrated for attribute optimization so that forecasting accuracy can be better. Based on the test results, it was found that SVR-PSO produced an RMSE value of 9.40.Keywords: sales forecasting; support vector regression; particle swarm optimization.Abstrak: Peramalan penjualan merupakan elemen penting dalam perencanaan strategi bisnis yang memiliki kemungkinan terjadi di masa depan. Tetapi bagaimana meningkatkan ketepatan proses peramalan saat ini masih menjadi tanda tanya besar bagi perusahaan. Penelitian ini menggunakan algoritme Support Vector Regression karena merupakan salah satu teknik peramalan yang dikategorikan baik dibandingkan dengan teknik yang lainnya. Algoritme Particle Swarm Optimization diintegrasikan untuk optimasi atribut agar akurasi peramalan dapat lebih baik. Berdasarkan hasil pengujian, didapatkan SVR-PSO menghasilkan nilai RMSE 9.40.Kata kunci : peramalan penjualan; support vector regression; particle swarm optimization.