Cardiovascular disease has been the major cause of mortality worldwide for last several decades. Diagnosis of heart disease through traditional approaches is a complex, time consuming and error prone process. To address this issue, several techniques have been proposed to automate the process of diagnosing the heart disease accurately in timely manner. However these techniques report limited accuracy of diagnosing the disease. In this paper SAFARI algorithm is used to diagnose the heart disease. Safari uses rule based approach i.e. it extracts rules from a dataset and uses the extracted rules for diagnosis. The various attribute values are first discretised into specific ranges, each range corresponds to a symbol. This results in a symbol table. Safari extracts rules from this symbol table. The approach has been thoroughly tested on the heart disease dataset publicly available on UCI machine learning repository. The results obtained using this approach are compared with the results of various techniques reported by other authors, a significant improvement was observed.
Read full abstract