Functionality Testing of Machine Learning Algorithms to Anticipate Life Expectancy of Stomach Cancer Patients

Abstract

Stomach Cancer is a strange development of cells that starts in the stomach. It can be called gastric cancer and can influence any stomach piece. All over the universe, malignant stomach development is the fifth -driving sort of disease and the third driving justification for death from threat. After being determined to have malignant growth, the doctor determines the patient's chances of survival and how long they can survive. The doctor usually estimates lifespan from his previous patient seeing experience; in some cases, estimation is wrong. But with the assistance of machine learning, it is possible to make this assumption very accurately. Typically individuals tackle these issues as regression issues. We have shown how the arrangement is conceivable with multiclass grouping. Moreover, the SEER data set guides us in our outing. Our created model can predict the sur-vival period of Stomach cancer patients. Exceptionally affected characteristics from SEER helped in the ML approaches. These high features feed to eight different classification algorithms: Extra tree, Random Forest, Bagging, Gradient Boost, LightGBM, XGBoost Decision tree, and HGB. The Extra Tree Classifier can predict the survival time with 97.27 % accuracy. These models will revolutionize the medical management of doctors.