Implementasi Support Vector Regression pada Prediksi Inflasi Indeks Harga Konsumen

Abstract

Inflation reflects an increase in the prices of these items as well as those used by the Indonesian government, especially Bank Indonesia, in determining monetary policy. An indicator that can be obtained by Bank Indonesia in measuring inflation is the Consumer Price Index. This study discusses inflation prediction using the SVR method. Inflation test data issued by Bank Indonesia. As a comparison material for the kernel used in the SVR method using two kernels, namely Linear and Radial Base Function. The error rate evaluation results show that linear kernels produce better values, with a MAPE rate of 8.70% and MSE of 0.0037