Acute Coronary Syndrome Prediction Using Data Mining Techniques- An Application

Last edited by

Quratt ul ain Siddique

Summary:

In this research paper data mining techniques are used to investigate the factors that are responsible for enhancing the risk of acute coronary syndrome. They have applied binary regression to factors that effecting the dependent variable. For the better performance of regression model in predicting coronary syndrome the reduction technique which is principle component analysis is used and applied. Based on results of data reduction, they have considered only 14 out of sixteen factors.

In this research paper logistic regression model is used to find the factors which are responsible for this Acute Coronary Syndrome (ACS). For the analysis of this problem data mining technique is used for the comparasion of the persons who have ACS or who don’t have.  

In this paper first data reduction techniques are applied that reduce the dimensions. After data reduction, the fourteen independent variables are age, gender, smoke, hypertension, family history, diabetics mellitus, fasting blood sugar, random blood sugar, cholesterol, streptokinase, blood pressure (systolic), blood pressure (diastolic), heart rate and pulse rate. After the calculation of corresponding significance of smoking which is “0” indicating that it has a high prevalence in the risk of ACS. The calculation of wald statistics indicates positive coefficients of HR, RBS, and BPS revealed that the risk of ACS increases with the increasing value of these factors.

The negative coefficients of BPd and PR revealed that the more the negative these values the more the increase in the risk of this disease. They observed that smoking is considered to be the worst cause of this Acute Coronary Sysndrome.