<p>This study presents a hybrid framework integrating Multi-Criteria Decision-Making (MCDM) and Machine Learning (ML) to optimize Wire Electrical Discharge Machining (WEDM) parameters for Inconel 925, a nickel-based superalloy critical to aerospace applications. A Taguchi L27 orthogonal array was employed to investigate the effects of six key parameters pulse-on time (Ton), pulse-off time (Toff), flushing pressure (FP), wire feed rate (WFR), wire tension (WT), and spark gap voltage (SGV) on material removal rate (MRR) and surface roughness (Ra). To ensure a robust and objective optimization, seven diverse MCDM methods (TOPSIS, VIKOR, MOORA, WPM, COPRAS, ARAS, COCOSO) were comparatively applied, utilizing entropy-derived weights (MRR: 0.62, Ra: 0.38). This multi MCDM approach consistently identified an optimal parameter configuration (Experiment 12) that achieved a 19% higher MRR and 15% lower Ra than the experimental average. Complementing the optimization, K-Nearest Neighbors (KNN) and Decision Tree Regression models were implemented to predict MRR and Ra. The KNN model demonstrated superior predictive accuracy with test R² values of 0.995 for MRR and 0.941 for Ra. Feature importance analysis from the ML models underscored Toff and Ton as the dominant factors, which aligns with mechanistic insights into thermal erosion and debris evacuation dynamics. This scalable framework provides a comprehensive methodology for enhancing precision manufacturing of difficult-to-machine alloys, supporting real-time process control and sustainable aerospace production.</p> Graphical Abstract: <p></p>