Machine Learning With Health Care: A perspective

ML “machine learning” is an ever-expanding research field with plenty of possibilities for study and implementation. Mr. James Collin stated at MIT that ML is the technology defining this decade, even though it has had a meagre effect on healthcare. Several fresh businesses in the ML industry are applying themselves earnestly on healthcare. Even google has jumped in to the race and it has designed a ML application for identification of cancer tumour on mammograms. To identify skin cancer, Stanford uses a Deep Learning algorithm. Around one trillion GB of data is getting generated per year by the USA health system. Various academic experts and scientists have worked out various characteristics and several factors of risk involved in of chronic illness. Additional data stands for more learning for machine, but for higher precision, these many features need a huge quantity samples. So if machines can harvest clinically greater risk oriented feature it would definitely be better. Precision is improved when the data in the form of exploratory data analysis and feature engineering is pre-processed. The multi-class classification might be capable of evaluating a patient’s different disease risk levels. In health care, correct identification of the percentage of diseased individuals “sensitivity” is a primary concern rather than correct identification of the percentage of healthy individuals “ specificity”. Our paper introduces one of ML’s super challenger and emergent application i.e. Healthcare. The careful and sympathetic relation with care providers will always be necessary for the patients. ML cannot remove this, but will become instrumental for health professionals in improving and strengthening on-going care. Our paper discusses the several established models of ML along with its applications in healthcare system. We also bring up directions for imparting more efficiency to ML model. In addition to this we have also discussed ML’s case study for Brest cancer.