Deep Learning-Driven Forecasting of Moroccan FDI: An LSTM-Based Approach

Abstract

This paper explores the transformative capabilities of Long Short-Term Memory (LSTM) models in macroeconomic prediction,| with a particular focus on forecasting the dynamics of foreign direct investment (FDI) within Morocco. Through meticulous analysis, our study showcases the remarkable proficiency of LSTM models in capturing the intricate trends and fluctuations inherent in FDI over time. Additionally, we explore the significance of features, scrutinize the residuals, compare real versus predicted values, and evaluate the robustness of the model. This comprehensive examination sheds light on the nuanced aspects of LSTM model performance and their potential implications for macroeconomic forecasting.