ISSN (Online) : 2456 - 0774

Email : ijasret@gmail.com

ISSN (Online) 2456 - 0774

 

Survey On Assess Co-Morbid Risk Of Diabetes Mellitus By UsingSplit And Merge Association Rule Summarization Techniques

Abstract

 A Diabetes is a life-threatening issue in 
modern health care domain. With the use of data mining 
techniques, diabetes factors and co morbid risk 
conditions associated with diabetes has found. In order 
to stifle the evolution of diabetes mellitus, applies 
distributed association rule mining and summarization 
techniques to electronic medical records. This helps to 
discover set of risk factors and co morbid conditions in 
distributed medical dataset using frequent item set 
mining. In general, association rule mining (ARM) 
generates bulky volume of data sets which need to 
summarize certain rules over medical record. This 
encompasses a novel approach to find the common 
factors which lead to high risks of diabetes and co 
morbid conditions associated with diabetes. This 
performs both association rule mining and association 
rule summarization techniques with improved 
classification algorithms. Exiting systems aim to apply 
association rule mining to electronic medical records to 
discover sets of risk factors and their corresponding 
subpopulations that represent patients at particularly 
high risk of developing diabetes. Given the high 
dimensionality of EMRs (Electronic Medical Records), 
association rule mining generates a very large set of rules 
which we need to summarize for easy clinical use. The 
existing system reviewed four association rule set 
summarization techniques and conducted a comparative 
evaluation to provide guidance regarding the diabetes 
risk prediction. In the field of medical domain, the 
prediction of diabetes and its Co-Morbid in earlier stage 
is important. We propose a set of methods to perform 
the Co-Morbid prediction. The propose technique 
named as SAM (Split and Merge), which is based on fast 
distributed quantitative association rule mining and rule 
filtering for prediction co morbid conditions associated 
with diabetes. SAM algorithm is used to discover the 
frequent data item sets and summarized data sets. In 
performance comparison of proposed SAM with existing 
BUS approach based on prediction efficiency SAM is 
better than BUS.
Keywords:- Association Rule Mining(ARM), Diabetes, Co-
Morbid, Risk Prediction, Electronic Medical Record (EMR), 
Split and Merge(SAM), Data Mining, Bottom-up 
Summerization (BUS).

Full Text PDF

IMPORTANT DATES

Submit Article (Vol. 5 Issue 3) 

B
efore 30 th March2020

Issue Publication   On 30 th March 2020