Inflation is a phenomenon that shows a continuous increase in the price of goods, which can cause a decline in the economic condition of a country. One of the indicators used to measure the inflation rate is the Consumer Price Index (CPI). By knowing the CPI value earlier, food prices can be controlled to be more stable. One method that can be used to predict CPI is Support Vector Regression (SVR), where this method is able to overcome linear and non-linear data conditions. This research aims to get the best prediction for CPI in South Kalimantan Province using CPI data for food groups in Tanjung, Banjarmasin, and Kotabaru in the 2014-2022 range. The best prediction results are obtained through the SVR method with Linear Kernel. The prediction error value measured through the MAPE value for Tanjung, Banjarmasin and Kotabaru is 0.77%, and . While the size of the meaning of the model measured through the coefficient of determination, respectively 0.8826, and . Based on these values, it is concluded that the prediction model formed is very good and feasible. The prediction results for the next 12 months show an increase, so that the government and related parties can formulate policies such as market operations and subsidy programs for the community.
Copyrights © 2024