| Sumario: | The optimization of regional development planning through the application of deep learning technology to predict economic trends is a strategic necessity in addressing the challenges of complex development dynamics and large-scale economic data. This study aims to examine the role of deep learning technology in enhancing the accuracy of economic indicator predictions such as inflation, Gross Regional Domestic Product (GRDP), and regional investment as a basis for decision-making at the local government level. Using a library research method through the review of indexed journals from Sinta, Google Scholar, and Scopus, the study finds that deep learning models such as Long Short-Term Memory (LSTM), ensemble learning, and Convolutional Neural Networks (CNN) with time series data fusion outperform traditional statistical methods, demonstrating lower prediction errors and greater adaptability to diverse and dynamic economic data patterns. In addition to improving data analysis efficiency and enabling simulations of policy scenarios based on data, key implementation challenges include the need for advanced technological infrastructure, skilled human resources, and effective data integration. The study recommends harnessing cloud computing, intensive HR training, and enforcing the One Data Indonesia policy to support the effective adoption of deep learning technology in regional development. The contribution of this technology is expected to strengthen evidence-based planning processes, create targeted policies, and support sustainable development that is responsive to global and local economic changes.
|
|---|