Cancer Death Cases Forecasting using Supervised Machine Learning
Sheshang Degadwala, Dhairya Vyas, Ankita Kothari, Urmila Khunt
Abstract
In India, like in the rest of the world, cancer is a major killer. This research objective is to predict cancer mortality in India, using supervised machine learning methods. Cancer mortality rates in India between 1990 and 2017 are provided by age group, gender, and region using data from the Global Burden of Disease Study. We employ three distinct supervised learning algorithms—linear regression, decision tree regression, and random forest regression—after performing data preprocessing, which includes missing value imputation and feature engineering. Using a variety of criteria, we analyze the effectiveness of these models and conclude that the random forest regression model is superior to the other two. The scope of research is provide a long-term prediction of cancer mortality in India using the best model so it will help health department to work on it. Our research has implications for policymakers and healthcare providers in India, where it may inform efforts to reduce cancer rates and improve cancer care.