Cervical cancer poses significant global health challenges, necessitating innovative treatment approaches. This study explores the integration of Support Vector Regression (SVR) in predicting radiation doses for cervical cancer radiotherapy, aiming to enhance the precision of Intensity Modulated Radiation Therapy (IMRT). We developed and validated an SVR model using datasets of 102 and 173 cervical cancer cases to predict dose distributions based on radiomic and dosiomic features. Our methodology involved pre-processing, feature extraction, data normalization, and rigorous training and testing of the model. Results indicated that larger, more diverse datasets significantly improve predictive accuracy, with the model demonstrating robust performance metrics. Notably, the study highlights the critical role of dataset size over stage uniformity in model reliability. These findings underscore the potential of machine learning in refining radiotherapy planning, reducing the workload on medical physicists, and improving patient outcomes. Future research should focus on expanding dataset sizes and enhancing model precision, particularly for challenging anatomical regions. This study paves the way for more personalized and effective radiotherapy planning, benefiting both clinicians and patients.