<![CDATA[Rice, as a strategic commodity in Indonesia, experiences significant fluctuations that affect economic stability and public welfare. This raises challenges in food policy planning, particularly in achieving accurate price forecasting. The research problem addressed is how to improve rice price forecasting accuracy using time series clustering capable of capturing complex interregional patterns. The study is limited to monthly rice price data over the past five years. To solve this problem, a hybrid model integrating Dynamic Time Warping (DTW) and Autoregressive Integrated Moving Average (ARIMA) is proposed. DTW measures similarity among time series patterns, while ARIMA constructs predictive models based on historical data. The objective is to produce more accurate and stable forecasts compared to single-method approaches. The urgency lies in its contribution to national food policy and food security. The novelty compared to previous studies is the integration of DTW-ARIMA within a time series clustering framework, rarely applied to food commodities in Indonesia. Experimental results indicate that the hybrid DTW-ARIMA model achieves lower mean absolute percentage error (MAPE) than conventional methods. The expected output is a predictive model that can be implemented as a data-driven decision support system for national food policy.]]>
Copyrights © 2026
