<![CDATA[In data mining, predictions are known to find knowledge about what will happen in the future. Predictions are usually made on time-series data. The Consumer Price Index (CPI) is an index value derived from daily consumer price data. The results of the CPI calculation are derived from observations of commodity prices at the household consumer level, which are carried out routinely on a daily, weekly, bi-weekly, and monthly basis. CPI prediction can be done using a data mining algorithm, namely Support Vector Regression (SVR). SVR is part of the Support Vector Machine algorithm that functions to solve regression cases. SVR is a reliable algorithm in the case of regression because it can handle data overfitting well. The data used as input in this paper comes from 34 food commodity prices, and the output data is obtained from the CPI value data. The food commodity price data used is from Surabaya City. The data period used is from 2014-2020. The results of the implementation of SVR with 4 kernels show that the Polynomial kernel has the best error rate with a MAPE value of 4.31%.]]>
Copyrights © 2022
