Predictive Modeling Techniques For Adverse Drug Reactions

The main goal of this paper is to predict adverse drug reactions from certain medications by leveraging deep learning and machine learning models to identify a relation (Adverse Drug Reaction Relation (ADRR)) between medications -specified in the Clinical Notes and Adverse Drug Events (ADE).The extraction of relations from clinical notes provides relevant information to identify the side effects of medications in postmarketing surveillance.Systems based on supervised learning solve relation extraction in clinical records, which requires rich features to learn effective models from the training data.Named Entity Recognition (NER) systems identify medical concepts in clinical notes like Medications and Indications.Contextual information is extracted via Deep Learning models and other different features are obtained from the ADRR.The proposed model improves the overall accuracy and the extraction of Adverse relations of the baseline, indicating the effectiveness of combining Deep Learning models and extensive feature engineering.This paper tries to depict the ease and accuracy which can be achieved with Deep Learning models as compared to traditional models.