<p>In modern industries, optimizing machining parameters is crucial for product quality, efficiency, and cost-effectiveness. Conventional methods often fail in the machining of advanced materials like single crystal tungsten. This study presents a hybrid Bayesian network model to optimise wire electrical discharge machining (WEDM) parameters for single crystal tungsten, combining numerical and experimental analyses to enhance the fabrication process. The Bayesian model determines optimal values and permissible ranges for key WEDM process parameters, including pulse-on time, pulse-off time, arc-off time, wire feed rate, and gap voltage. Numerical simulation is carried out by the finite element method, indicating that single crystal tungsten has an ultimate tensile strength of 850&#xa0;MPa with limited ductility, confirming its brittle nature. Subsequently, single crystal tungsten is fabricated under the optimised WEDM conditions determined by a hybrid Bayesian model and the numerically obtained ultimate compressive strength. Furthermore, comprehensive experimental analyses, such as X-ray diffraction, Laue diffraction, scanning electron microscopy, and X-ray fluorescence, are performed both before and after fabrication of a single crystal tungsten specimen to examine its crystallinity, microstructure, and compositional integrity. Overall, the experimental analysis demonstrates that optimised WEDM parameters, determined by a hybrid Bayesian network, are effective for fabricating single crystal tungsten and anticipated for critical technological applications.</p> Graphical abstract <p></p>