PRIOR PREDICTION AND IMPEDIMENT OF CANCER USING MACHINE LEARNING PROCESS

In today’s life rapid changes occur drastically in climatic, food habits and heredity of gene mutations. Human health factor is causing the growth of death rate worldwide. Among the various dead causes, caner is the leading disease worldwide. Early prevention and detection of cancer will play a vital role in decreasing the deaths caused by cancer disease. Present work proposes a hybrid method of multi-layered which combines decision tree classification and clustering technique for building a prediction and cancer risk system. The proposed work will predict, breast, lungs, cervix, oral, blood and stomach cancers and is a user interface system which reduces cost and time saving. The present work is based on machine learning method which uses clustering, classification and prediction in identifying the probability of cancer in human being. The datasets collected are preprocess, feeded to the classifier which yields related patterns based on decision making algorithm.The data generated is clustered using k-means algorithm of clustering in separating non-cancer with that of cancer patient. Further it is found that the clusters of cancer are divided into 6 categories. The early prediction system developed analyzed various levels of risk and helped in diagnosis. Present research help in detection the predisposition of person early canceric conditions.