Electronic medical record (EMR) data are becoming increasingly important but most implementations fails to support analysis of aggregated data. As a result, mining of datasets becoming a difficult task with text based classification. In this work, we present an integrated classification model that combines a feature extraction algorithm with a classifier. The feature extraction algorithm operates in such that it examines the unclustered documents in an automated way after the extraction of relevant features. Further, the classification algorithm classifies the extracted features to provide relevant class from the given input samples. In this work, we present an integrated classification model that combines a feature extraction algorithm (Reinforcement Learning - RIL) with a classifier (Naïve Bayes with Decision Trees NB-DT). The feature extraction algorithm using Reinforcement Learning operates in such that it examines the unclustered documents in an automated way after the extraction of relevant features. Further, the classification algorithm using NB with DT classifies the extracted features to provide relevant class from the given input samples. The results shows that the proposed method has higher classification rate than the existing methods.
Volume 12 | Issue 6
Pages: 2254-2261
DOI: 10.5373/JARDCS/V12I6/S20201184