Development of Predictive model for survival of Paediatric HIV/AID Patients in South Western Nigeria using Data Mining Techniques.
Loading...
Date
2014
Authors
Journal Title
Journal ISSN
Volume Title
Publisher
Obafemi Awolowo University
Abstract
The study identified survival variables for HIV/AIDS paediatric patients, developed predictive models for determining the survival of the patients who were receiving antiretroviral drug in the Southwestern Nigeria based on identified variables, compared the developed models and validate the models with the historical data. This was with a view to choosing the most efficient model for the prediction of the survival of Peadiatric HIV/AIDS patients
Interviews were conducted with the virologists and Peadiatrician at two health institutions from the study area in order to identify survival variables for HIV/AIDS Paediatric patients. Paediatric HIV/AIDS patients’ data (216) were also collected from two health institutions, preprocessed and the 10-fold cross validation technique was used to partition the datasets into training and testing data. Predictive models were developed using three (3) supervised learning techniques (Naïve Bayes’ classifiers, decision trees and the multi-layer perception (MLP)) and the Waikato Environment for Knowledge Analysis (WEKA) was used to simulate the models in which CD4 count, Viral Load, Opportunistic infections and Nutritional status were used as the independent variables for the prediction.
The result showed that all the three techniques (Naïve Bayes’ classifiers, decision trees and the multi-layer perception (MLP)) were suitable in carrying out the task of forecasting the survival of Paediatric HIV/AIDS patients so that each patient can know their status at every point in time. The decision trees model has an accuracy of 99.07% (214 correct classifications out of 216), 0.0183 mean absolute error rate, 0.0962 root mean square error and 3.69% relative absolute error. The Receiver Operating Characteristics (ROC) area for the model was also 0.993 showing that the level of bias was very low (0.007), Naïve Baye’s model has an accuracy of 81.02% (175 correct classifications out of 216), the mean absolute error rate was 0.2025, 0.2920 for the root mean square error and 40.92% for the relative absolute error. The ROC area for the model was also 0.993 showing that the level of bias was very low (0.007) and multilayer perception model has an accuracy of 99.07% (214 correct classifications out of 216), the mean absolute error rate was 0.022, 0.0962 for the root mean square error and 4.48% for the relative absolute error. The ROC area for the model was also 0.992 showing that the level of bias was also very low (0.008). The result of the three models showed that the decision tree model was the most efficient of all the three models with an accuracy of 99.07% (214 correct classifications out of 216). The validation was done by comparing the three developed models with the historical data from the two selected health institutions which has a catchment of patients from the other parts of the South western Nigeria.
The study concluded that the decision trees technique is viable in predicting survival among HIV/AIDS patients in Southwestern Nigeria and that the prediction of poor prognosis for survival may mean that patients, relations and care givers will need to work assiduously and see how the negative survival can be changed.
Description
xv,151P
Keywords
HIV/AIDS, Paediatric Patients, Virologist, Peadiatrician, SouthWestern, Nigeria
Citation
Agbelusi,.O(2014)Development of Predictive model for survival of Paediatric HIV/AID patients in south western nigeria using data mining techniques.Obafemi Awolowo University.