<![CDATA[This research aims to implement the Beale-Powell Restarts algorithm in predicting the development of oil and non-oil exports in Indonesia. With the increasing importance of international trade for the economic growth of a country, accurate understanding of export trends becomes crucial for decision-making at the policy level. The data used in this study originates from the value of oil and non-oil exports (in Million US$) in Indonesia obtained from the customs documents of the Directorate General of Customs and Excise (PEB and PIB). The implementation method of the Beale-Powell Restarts algorithm is focused on analyzing and forecasting export development trends. This algorithm is known for its ability to address convergence issues commonly encountered in nonlinear optimization. By applying this algorithm, the research aims to improve the accuracy and precision of predictions, providing valuable insights for economic planning and trade strategy development in Indonesia. The study also includes a comparison of the performance of several different models in prediction, with various models such as 8-5-1, 8-10-1, 8-15-1, 8-20-1, and 8-25-1 being evaluated. The research findings indicate that the best model is 8-5-1, which has the lowest testing Mean Squared Error (MSE) value of 0.00735820154, affirming that the use of the Beale-Powell Restarts algorithm yields better results in predicting the development of oil and non-oil exports in Indonesia compared to other models. It is hoped that the implementation of the Beale-Powell Restarts algorithm will make a significant contribution in assisting stakeholders in formulating more effective and sustainable trade policies to advance the Indonesian economy.]]>
Copyrights © 2024
