Abstract
This research challenges conventional economic expectations by delving into the intricate dynamics of economic relationships. Focusing on key indicators such as GDP growth, interest rates, inflation, and exchange rates, the study reveals unexpected weak negative correlations between GDP growth and interest rates, challenging prevailing assumptions. Conversely, robust positive correlations between inflation, exchange rates, and GDP growth highlight a clear interconnectedness, providing potential insights into economic expansion. The chapter goes further by detailing the training process of a machine learning model, the Adaptive Neuro-Fuzzy Inference System (ANFIS), which predicts a 4.2225% GDP growth in 2024 after 10 epochs. Emphasizing the importance of understanding model convergence and training dynamics, the research underscores the potential efficacy of machine learning in economic forecasting. ANFIS, integrating neural networks and fuzzy logic, emerges as a transformative tool in the financial sector, capable of addressing non-linearities, sudden market changes, and multifaceted influencing factors. Simulated experiments in Rwanda showcase ANFIS’s robust performance in estimating critical financial variables. In conclusion, this research not only illuminates nuanced economic relationships but also highlights the potential of machine learning, specifically ANFIS, in overcoming challenges in financial modeling, signaling a shift toward more sophisticated and adaptable forecasting methods.