A Framework To Support Management Of HIVAIDS Using K-Means And Random Forest Algorithm
Journal: International Journal of Scientific & Technology Research (Vol.6, No. 6)Publication Date: 2017-06-15
Authors : Gladys Iseu; Waweru Mwangi; Michael Kimwele;
Page : 61-68
Keywords : Clustering; Classification; K-Means; Random Forest; Data Mining; Big Data;
Abstract
Healthcare industry generates large amounts of complex data about patients hospital resources disease management electronic patient records and medical devices among others. The availability of these huge amounts of medical data creates a need for powerful mining tools to support health care professionals in diagnosis treatment and management of HIVAIDS. Several data mining techniques have been used in management of different data sets. Data mining techniques have been categorized into regression algorithms segmentation algorithms association algorithms sequence analysis algorithms and classification algorithms. In the medical field there has not been a specific study that has incorporated two or more data mining algorithms hence limiting decision making levels by medical practitioners. This study identified the extent to which K-means algorithm cluster patient characteristics it has also evaluated the extent to which random forest algorithm can classify the data for informed decision making as well as design a framework to support medical decision making in the treatment of HIVAIDS related diseases in Kenya. The paper further used random forest classification algorithm to compute proximities between pairs of cases that can be used in clustering locating outliers or by scaling to give interesting views of the data.
Other Latest Articles
- Penetration Testing In System Administration
- The Banks Rating Analysis The Differences Between The Regional Development Banks And Non-Foreign Exchange Commercial Banks In Indonesia
- Modelling And Dynamic Analysis Of Catenary Maintenance Vehicle By AdamsVI-Rail
- Tailoring The Conducting Polymers PPY And PANI With Ionic Liquid BMIMBr For Enhanced Electrochromic Properties
- Software Testing And Defect Analysis Using Soft Computing Concepts
Last modified: 2017-10-22 19:55:28