<p>Carbon Capture and Utilization are regarded as an effective method for decarbonizing hard-to-abate industries, supporting Net-Zero goals. CO₂ removal by absorption uses chemical solvents either in a conventional absorption-regeneration column or in a Rotating Packed Bed (RPB). RPB, a process-intensified technique, has gained acceptance due to its superior mass transfer. For industrial scale-up purposes, modelling and predicting CO₂ absorption in RPB is crucial to identify factors contributing to enhanced absorption efficiency. The machine learning (ML) approach is currently a popular method in this context. In this work, the aim is to employ different ML techniques, such as XGBoost, Random Forest, GPR, BPNN, and RBFNN, to predict the CO₂ composition in the gas outlet with changing inputs such as gas composition, flow rates, temperature, pressure, and rotational speed of RPB. The data, collected from literature, comprise approximately 65 data points that utilize MEA solvent for CO₂ absorption in RPB. The performance of ML models has been assessed using graphical analysis along with statistical measures such as R², MAE, AAD%, and RMSE. The results have also been compared with a mathematical model reported in the literature. The outcomes of the work suggest that out of all five models, GPR demonstrates the best accuracy with an R² value of 0.98 and an AAD% of 1.07%. The results have been further utilized in SHAP analysis to identify which input features primarily affect the predictions, and optimization has been integrated to determine the maximum absorption efficiency with optimal operating parameters. The modelling results provide a framework for performance analysis of CO₂ capture using RPB, combining interpretability, accuracy, and optimization for industrial plant implementation.</p>