RETRACTED: Analysis and Prediction of Adverse Reaction of Drugs with Machine Learning Models for Tracking the Severity (Retracted Article)

Reaction of drugs can be adverse or favourable when it is tested against the disease in the diagnosis process. There are numerous tests for drugs in different perspectives, to validate the effects and side effects over human beings. Evaluation of drug and their reactions is analysed through stringent and sequential processes with timely data monitoring. Hence, the proposed method predicts the adverse reaction of new drugs. Medical records have been digitized and maintained as electronic health records (EHR) or electronic medical records (EMR). Drug re-purposing and the side effects of drugs can be predicted through an automated framework for achieving better accuracy. Complex chemical structures and results of post-consumption detection of effects can be eliminated by this advanced architecture. The proposed work comprises a novel technique to address the limitations of the existing data mining techniques to predict the severity of reactions when a drug is consumed. The proportion of the drug is analysed and estimated for a precision using a statistical hypothesis test to map the reaction and proportion of the chemical used. The statistics extracted from various data mining techniques are fed into machine learning algorithms implemented with decision tree and support vector machines to derive the adverse drug reaction. The intensity of the drug reaction depends on other parameters such as age, gender, weight and other demographic information available in the electronic records. The proposed work achieved an accuracy weight of 91.8 per cent in predicting fatality as the result of the adverse drug reaction.