With an ensemble learning-supportive model for multiple-class differentiating of arrhythmia disorder that is a hybrid of biometrical and clinical related data. The system uses several supervised learning mechanism to identify complex data patterns and relation that are not revealed from the corresponding of individually from the classifiers. The ensemble strategy not only provides the robustness but also gains better accuracy in the classification of various arrhythmia types through the exchange of the complementary strengths of different algorithms. The model will made up of unseen described the input preprocessing and related feature fusion methods that make it easier to use several information sources and, it will, allow for better generalization and data consistency. It also, the schema will make a customized way of handling class misbalancing and making the prediction process more accurate manner for the minority factors that are typically less given in medical related datasets. The suggested methodology outlines an end-to-end streamlined from data integration to differentiate and which allows for a scalable and flexible clinical based decision support system for cardiac process.