This paper introduces a integrated machine learning model designed to predict stroke risk by analyzing various contributing factors. The model was rigorously trained and validated on an imbalanced dataset, encompassing numerous features associated with stroke risk. To enhance the model's performance, we applied hybrid sampling techniques to address data imbalance and used Grid Search to precisely identify the optimal parameters for our five foundational models. Among these, the Integrated Classification Method achieved remarkable precision, recall, accuracy, and F1-score, ranging from 94% to 95%. This high level of performance underscores the model's robustness and its potential to significantly aid clinical decision-making in stroke risk prediction.