Classification And Regression Tree In Prediction Of Survival Of AIDS Patients

Main Article Content

Sameem Abdul Kareem
S. Raviraja
Namir A Awadh
Adeeba Kamaruzaman
Annapurni Kajindran

Abstract

Over the years, the advancement in computing technology, the reliability of computers, coupled with the development of easy-to-use but nevertheless sophisticated software has led to significant changes in the way that data are collected and analyzed. Computations has shifted from off-site main frames, dependent on highly trained operators and located in special rooms accessible only to certain authorised staff, to the more accessible desktop and laptop computers. This accessibility has resulted in an increasing number of researches in data mining in which hidden predictive information are extracted from large databases, using techniques from database research, artificial intelligence and statistics, to a wide variety of domains such as finance, manufacturing and medicine. In this research we describe our experiments on the application of Classification And Regression Tree (CART) to predict the survival of AIDS. CART builds classification and regression trees for predicting continuous dependent variables and categorical or predictor variables, and by predicting the most likely value of the dependent variable. In this paper, a total of 998 patients who had been diagnosed with AIDS were grouped according to prognosis by CART. We found that CART were able to predict the survival of AIDS with an accuracy of 60-93% based on selected dependent variables, validated using Receiver Operating Characteristics (ROC). This could be useful in determining potential treatment methods and monitoring the progress of treatment for AIDS patients.

Downloads

Download data is not yet available.

Article Details

How to Cite
Abdul Kareem, S., Raviraja, S., A Awadh, N., Kamaruzaman, A., & Kajindran, A. (2017). Classification And Regression Tree In Prediction Of Survival Of AIDS Patients. Malaysian Journal of Computer Science, 23(3), 153–165. Retrieved from https://jml.um.edu.my/index.php/MJCS/article/view/6408
Section
Articles

Most read articles by the same author(s)