Performance Analysis of Learning Models On Medical Documents
Author(s):
Vanitha Guda, Manish Golla, Akhilesh Datta
Keywords:
Classifier’s Accuracy, Document classification, Feature Selection, Learning Models, Medical Documents, Text Classification
Abstract
With the exponential growth of online text, Text Classification domain becomes the major field of Natural language Processing and Machine learning. In this context Medical Document Classification is one of the popular research problem to analyze the high dimensionality features of medical data. Our Study considered various learning models and their performances over the medical documents and we considered OSUMED is one of the popular datasets containing MEDLINE documents as multi-labelled documents. Choosing a high accuracy classifier for text classification is still a challenging task for many of the practitioners. Our work aims to find the efficiency in classifiers and comparing the accuracy in classifying medical documents with well-known classifiers Naïve Bayes, Decision Tree, Support Vector Machine (Linear) and Stochastic Gradient Descent (SGDC). The performance of a feature selection method namely Univariate Feature Selection is analyzed using pattern classifiers namely Naïve Bayes, Decision Tree, Support Vector Machine (Linear) and SGDC and the obtained experimental results shows that the combination of Univariate Feature Selector and Support Vector Machines classifier gives more accurate results in most cases than the others.
Article Details
Unique Paper ID: 146500
Publication Volume & Issue: Volume 4, Issue 12
Page(s): 822 - 828
Article Preview & Download
Share This Article
Join our RMS
Conference Alert
NCSEM 2024
National Conference on Sustainable Engineering and Management - 2024