Disease Prediction and Risk Analysis using Classification Algorithms

Mixymol V.K.


Data mining techniques helps to extract important information from a large set of data set which may be widely spread and also helps to find correlation between these data sets. A huge amount of medical data is available at present which are widely spread. Currently a large number of tests are done to reach to a conclusion and this large number of test results may lead to lots of confusion and may lead to complication of the diagnosis process. The major issue with medical diagnosis is to reach to a correct diagnosis to take the correct decision. In this context, machine learning with the help of artificial intelligent algorithms can make use of decision making for rule generation process in healthcare data.
Objective of this work is to analyze the performances of disease prediction using various classification techniques available on WEKA data mining tool. Data set for the present work is collected from UCI repository. Data set is collected for Dermatology, Hepatitis and Heart disease. Three classification techniques, Naïve Bayesian, Nearest Neighbor and Reptree, are considered for the performance analysis since it is found that these three algorithms are commonly used in disease prediction in many studies. With the future development of information and communication technologies, data mining will achieve its full potential in the discovery of knowledge hidden in the medical data.


Data mining, Machine learning, Medical data, Data mining tool, Disease prediction, Performance Analysis.

Full Text:


DOI: https://doi.org/10.26483/ijarcs.v8i5.4094


  • There are currently no refbacks.

Copyright (c) 2017 International Journal of Advanced Research in Computer Science