<![CDATA[PT XYZ does not have a tested evaluation metric to measure the accuracy of procurement prediction models. This research aims to find a suitable metric by comparing two methods: Autoregressive Integrated Moving Average (ARIMA) and Fuzzy Time Series (FTS). Both methods were chosen based on the ability of ARIMA to handle time patterns and trends, and the flexibility of FTS in dealing with uncertainty in procurement values. The research uses Root Mean Squared Error (RMSE) values to measure prediction accuracy. The ARIMA implementation analyzes historical data to forecast patterns, while FTS accommodates fluctuations and uncertainties, allowing for more adaptive and accurate predictions. The analysis results show that the ARIMA model has an AIC value of 3953.57 and a residual value of 3351745.26, while FTS has a residual of -224.79. The RMSE evaluation shows the ARIMA value of 3351745.30 and FTS of 224793895.00. The predicted value of ARIMA is 440,326,255, while FTS is 668,471,895. Based on these results, FTS shows superior prediction accuracy compared to ARIMA. FTS is recommended to be implemented at PT XYZ due to its ability to effectively manage uncertainty and fluctuations in procurement values, enabling more accurate strategic decisions. Further research is needed to understand the factors that influence the performance difference between these two methods.]]>
Copyrights © 2024
