Cancer Detection and Classification using Random Forest, NN and XGBoost Algorithm of Machine Learning

Abstract

Healthcare is the topmost high priority subject matter spoken about. Detailed diagnosis of the disease is required to give individuals the required attention in time. Advancements in healthcare using ML technologies have been beneficial for the health component of exploration and early diagnosis. The extremely high number of new methodologies permits new pathways favorable for the society. The foremost period is the most crucial which when detected early can be easily curable. Branches of machine learning which includes   deep learning whose applications have lately gained more importance in medical text and image research due to their benefits and success. Step has been taken to assess studies on ML and DL methods used to mimic different types of cancer using these three categories including the intent of the prediction, approach of prediction, and data instances. For this, we suggest systematic examination of different human cancerous diseases by applying techniques such as NN, XGBoost, Random Forest to make important predictions and help in decision-making. By using this methodology, the system proposed in this work is able to achieve 99.45% accuracy and 99.95% AUC in detecting whether the patient has cancer or not and achieved 93.94% accuracy in classifying the cancer types. Our methodology has been successful in solving the patients problem with recommendable results.