<p>Designing aerospace vehicles demands accurate prediction of various aerodynamic coefficients, but Computational Fluid Dynamic (CFD) based simulations are often computationally expensive and time-consuming. As demand for more efficient and expedited design processes in aerospace engineering rises, alternative solutions are in high demand. This study proposes a novel approach for predicting the drag coefficient (Cd) of a rocket nosecone using several Machine Learning (ML) techniques trained on CFD-generated data. Four regression models—Random Forest, Decision Tree, Support Vector Regression (SVR), and Polynomial Regression—are evaluated using 200 simulation-based data points across Mach 0 to 1. The models were subsequently compared and assessed using important performance metrics, including mean absolute error (MAE), coefficient of determination (<i>R</i>²), explained variance score (EVS), and root mean square error (RMSE). Results show that the Random Forest model achieves the highest predictive accuracy with an <i>R</i>² value of 0.9863 and lowest RMSE of 0.0031. This method significantly reduces analysis time while maintaining high reliability, offering a scalable alternative to conventional CFD for preliminary aerodynamic design. The findings establish a pathway for incorporating ML tools into real-world aerospace workflows to optimize performance and resource use.</p>