Knowledge Mining and Trifle Management for Predicting Heart Disease

Mr.Bensujin Bensujin, Dr.Kezi selva vijila, Mrs.Cynthia Hubert


Heart Disease (HD) is the foremost cause of death worldwide; the World Health Organization(WHO) estimates that globally 17.3 million people died from Heart Disease in 2008, representing 30% of global deaths. The prediction of heart disease from various inputs and raw data is a multi-layered problem, which is not free from false assumptions. It is essential to extract the valuable information to spot the risk of heart disease. This papers details about an online expert system to identify the presence of heart disease in any human by the use of risk factors. The risk factors are the life style habits or the behavioural characteristics present in the analysis individual. The OES system calculates an OES scoring based on the risk factors and the decision is taken based on the score by the artificial neural networks. The risk factors and the inputs to the system is given in the form of trifles. The obtained input datas are pre-processed and clustered using K-means clustering algorithm. The fine tuned data set is converted into trifles and passed to the Artificial Neural Network(ANN) for the identification of heart disease. A Multi Layered feed forward network is used for the decision making and the Back Propogation(BP) algorithm is used to train the network and adjust the weight of the nodes in the neural network. The OES System enables significant knowledge, e.g. patterns, relationships between medical factors related to heart disease, to be established. OES is Web-based, user-friendly, scalable, reliable and expandable system and it consist of a simple database designed by the Microsoft tool and the system is implemented on the .NET platform using ASP.


Keywords: Trifles, Heart Disease, Back Propogation, risk factors, K-means clustering, WHO.

Full Text:




  • There are currently no refbacks.

Copyright (c) 2016 International Journal of Advanced Research in Computer Science