Archives

Heterogeneous Clustering Network Chronic Kidney Disease Progression Mining (HCNCKPM)


Diwakar Bhardwaj and Anjani Rai
Abstract

Around the world, billions of dollars have been lost due to the fraud in healthcare. This is specifically high in case of a disease called chronic diseases. Complex relationships like time-gap and disease pattern occurrence are not considered in the existing approaches. The healthcare insurance fraud detection can be done with the help of various stages of same chronic disease. Healthcare cost also reduced by this. Recently Heterogeneous Network-based Chronic Disease Progression Mining (HNCDPM) method is introduced for predict of chronic diseases. HNCDPM uses a cosine similarity function it doesn‟t enhances the efficiency for increased number of the samples, so some advanced clustering methods are required to solve this issue. In this work, Average bound Crossover based Fuzzy C Means (ACFCM) Clustering is proposed instead of cosine similarity function for measuring the similarity between the temporal graphs and frequent subgraph. The current understanding about chronic disease progression can be improved by proposed Heterogeneous Clustering Network Chronic Kidney disease Progression Mining (HCNCKPM) in this work. Relations among different chronic diseases is indicated by pattern between various stages of various chronic diseases. Clinical path of chronic disease is indicated by pattern between various stages of same chronic disease. Finally the results are measured with respect to F-measure, recall and precision with CKD database from machine learning repository

Volume 11 | 11-Special Issue

Pages: 215-222

DOI: 10.5373/JARDCS/V11SP11/20192950