ADVERSE DRUG REACTION USING LEVENBERG-MARQUARDT BACKPROPAGATION ALGORITHM

Abstract

Accurate clinical studies are needed before a drug enters into the market, but it is impossible to forecast all side effects for the approved drug. So, a database is created by collecting the information by the patient, pharmacy and doctor report about adverse drug reaction. The collected data are analysed through sentimental analysis and gives positive, neutral and negative percentage of the text. The percentage values are evaluated by our proposed neural network such as Levenberg-Marquardt backpropagation algorithm. The results show the effective way to identify the harmfulness of the drug and the adverse drug reaction.