This study presents an innovative approach combining numerical simulations with experimental data to improve the accuracy of tool wear prediction. A hybrid modeling strategy is employed, integrating physics-based finite element method with data-driven machine learning techniques. Tool life experiments in turning operations were conducted, and a tool wear state-dependent finite element model was developed alongside an acoustic emission-based extreme gradient boosting regression model. Cutting forces calculated through the finite element model were integrated into the machine learning model to enhance predictive performance. The results show that incorporating simulated process data significantly improves wear prediction capabilities and accuracy compared to purely data-driven models. This demonstrates the potential of hybrid modeling approaches, so called grey-box, to bridge the gap between physical process understanding and machine learning predictions, minimizing the need for extensive experimental data collection. Furthermore, this approach reduces the dependency on expensive measurement technologies by substituting real measurement data with simulated data. By leveraging these advancements, this research contributes to the development of a robust and reliable tool wear prediction system, which not only improves manufacturing efficiency but also reduces operational costs in the future.