Exploring Significant Heart Disease Factors based on Semi Supervised Learning Algorithms

摘要

Heart Disease is one of the leading diseases that causes enormous loss of lives all over the world. There are happened many works to diagnosis heart disease. In this paper, we are considered some unusual approaches to find out significant factors of heart diseases. There are considered two heart disease data (Cleveland & Hungarian) and both of them are divided into 33%, 65% and 100% data. Values of different range of individual attributes in these data are determined to find out relevant factors of this disease. Then, different semi supervised learning algorithms such as Collective Wrapper, Filtered Collective and Yet Another Semi Supervised Idea are used to analyze heart disease data. There are considered some metrics of these classifiers like accuracy, f-measure and area under ROC to justify individual classifiers and specify the best semi supervised learning algorithm. This algorithm is explored significant and irrelevant factors of heart disease by removing attributes one after another sequentially and observing the outcomes of classification. Experiment results on two real data demonstrates the effectiveness and efficiency of our analysis.