<![CDATA[Cooking oil is a strategic staple food commodity in Indonesia. The high consumption of cooking oil every year is accompanied by increasing demand from the market, causing the price of cooking oil to increase every year. The purpose of this study is to group provinces in Indonesia based on time series patterns of cooking oil prices and evaluate group-level forecasts. The clustering algorithm used is Agglomerative Hierarchical Clustering (AHC) with dynamic time warping (DTW, autocorrelation function (ACF), and Euclidean distance similarity measures. The optimal algorithm and number of clusters is selected based on the cophenetic correlation coefficient and silhouette coefficient. Then, each cluster that is formed will be modeled Autoregressive Integrated Moving Average (ARIMA) with the Auto ARIMA approach. Forecasting evaluation using Mean Absolute Percentage Error (MAPE). This study shows that the optimal algorithm chosen is the average linkage with a euclidean distance’s similarity measures. Cophenetic correlation and silhouette coefficient obtained respectively 0.8281 and 0.4296. The MAPE values obtained were 0.3399 and 0.0793 respectively.]]>
Copyrights © 2022
