Prediction Of Long Term Living Donor Kidney Graft Outcome: Comparison Between Different Machine Learning Methods
Journal Title: INTERNATIONAL JOURNAL OF COMPUTERS & TECHNOLOGY - Year 2014, Vol 14, Issue 2
Abstract
Predicting the outcome of a graft transplant with high level of accuracy is a challenging task In medical fields and Data Mining has a great role to answer the challenge. The goal of this study is to compare the performances and features of data mining technique namely Decision Tree , Rule Based Classifiers with Compare to Logistic Regression as a standard statistical data mining method to predict the outcome of kidney transplants over a 5-year horizon. The dataset was compiled from the Urology and Nephrology Center (UNC), Mansoura, Egypt. classifiers were developed using the Weka machine learning software workbench by applying Rule Based Classifiers (RIPPER, DTNB),Decision Tree Classifiers (BF,J48 ) and Logistic Regression. Further from Experimental Results, it has been found that Decision Tree and Rule Based classifiers are providing improved Accuracy and interpretable models compared to other Classifier.
Authors and Affiliations
Maha Fouad, Dr. Mahmoud M. Abd ellatif, Prof. Mohamed Hagag, Dr. Ahmed Akl
Keywords
Related Articles
- EP ID EP650616
- DOI 10.24297/ijct.v14i2.2066
- Views 82
- Downloads 0
A Survey on SmartPhone Honeypot
Smartphones are becoming  a dominant form of mobile computing in the world. The Smartphone, as a platform, blends a traditional general computing platform with a specialized mobile phone platform. The general computin...
Calcination and alkali metal modification of alumina as solid stationary phase in GC
Alumina has unique textural properties and functionalized surface. The single or multi-treatments of alumina with calcination and alkali metal treatment (NaCl) are of prime one importance not only for determining the str...
Integration of GPS with Digital devices and Interactive objects for public safety.
Abstract: The paper demonstrated the role of GPS technology in law enforcement and public safety application areas. It presents the design, implem...
University Course Timetabling using Multi-population Genetic Algorithm Guided with Local Search and Fuzzy Logic
Problem of courses timetabling is a time consuming and demanding issues in any education environment that they are involved in every semester. The main aim of timetabling problem is the allocation of a number of courses...
Implementing Clustering Based Approach for Evaluation of Success of Software Reuse using K-means algorithm
A great deal of research over the past several years has been devoted to the development of methodologies to create reusable software components and component libraries. But the issue of how to find the contribution of t...