Efficient Heart Disease Prediction System using K-Nearest Neighbor Classification Technique

The healthcare industry gathers gigantic volumes of healthcare data, which can be turned into useful information by using different Data mining techniques. This knowledge can then be useful in many different decision making environments like Finance, Marketing, Healthcare sector. In health care industry, cardiovascular disease is one of the most significant but difficult task that needs to be done very swiftly, efficiently and the correct automation is desirable. In this regard, different Heart Disease Prediction systems have been reviewed and the work done previously have applied different numbers of medical parameters and risk factors with different data mining techniques. Accuracy claimed in each of those papers depends upon the number of parameters and data mining techniques applied. However, the number of the parameters and datasets to perform experiments was very small to claim high accuracy. We employed K-nearest neighbor classifier to achieve an accuracy of approximately 80% by using 14 attributes. Moreover, a number of predominant classifiers have been evaluated to highlight the supremacy of k-NN classifier based heart disease prediction system.