Predictive Model for Estimating 5-Year Survivability of Osteosarcoma Patients Using Ensemble Learning

A tumour is identified as an uncontrolled growth of any particular type of tissue in body. They can further be classified as malignant tumours, which are the cancerous ones and benign tumours, which are the non-cancerous ones. Osteosarcoma is a malignant bone tumour, usually caused in children and youths. In general, the primary bone cancer accounts for the 1% of the novel cancer cases diagnosed every year, but the secondary bone cancer is more common in nature. This research has established a model, based on Stacking ensemble learning technique for making the final prediction about the next 5 years survivability of Osteosarcoma patients. Total 3730 instances of Osteosarcoma patients were taken from SEER database as input for this purpose. Initially classification techniques like Support Vector classifier and Bagging were applied independently. Later, the Stacking technique was used to achieve the absolute predictions with the validation time accuracy of 92.08%. This research further includes the comparison among the f1, accuracy, precision, and recall scores registered by the various applied classifiers to state the validity of the proposed model. Further, the performance of the proposed approach is found to be better than that of other existing approaches in the same class.