UCLA

This thesis investigates the application of machine learning models—Linear Regression, Support Vector Machine (SVM), and Random Forest—in predicting bond market trends, a critical area of financial forecasting. Using data from 2019 to 2023 sourced from the Federal Reserve Economic Data (FRED), key economic indicators such as the 10-year Treasury yield, inflation (CPI), unemployment rate, and Federal Funds Rate were analyzed. Random Forest demonstrated superior performance, achieving the highest predictive accuracy and lowest error metrics. The research also identifies inflation and the Federal Funds Rate as the most influential variables, emphasizing the capacity of machine learning to capture nonlinear relationships and enhance data-driven decision-making. While promising, the study acknowledges limitations such as a restricted dataset timeframe and model complexity. Future research directions include exploring advanced deep learning techniques, kernel transformations for SVM, and expanding the feature set to include geopolitical and sentiment-based variables. This study contributes to financial analytics by showcasing how machine learning models can improve forecasting accuracy and decision-making in bond market analysis.