Heart disease is the primary cause of death nowadays. Treatments of heart disease patients have been advanced, for example with machine-to-machine (M2M) technology to enable remote patient monitoring. To use M2M to take care remote heart disease patient, his/her medical condition should be measured periodically at home. Thus, it is difficult to perform complex tests which need physicians to help. Meanwhile, heart disease can be predicted by analysing some of patient's health parameters. With help of data mining techniques, heart disease prediction can be improved. There are some algorithms that have been used for this purpose like Naive Bayes, Decision Tree, and k-Nearest Neighbor (KNN). This study aims to use data mining techniques in heart disease prediction, with simplifying parameters to be used, so they can be used in M2M remote patient monitoring purpose. KNN is used with parameter weighting method to improve accuracy. Only 8 parameters are used (out of 13 parameters recommended), since they are simple and instant parameters that can be measured at home. The result shows that the accuracy of these 8 parameters using KNN algorithm are good enough, comparing to 13 parameters with KNN, or even other algorithms like Naive Bayes and Decision Tree.
|Number of pages||7|
|Journal||Journal of Telecommunication, Electronic and Computer Engineering|
|Publication status||Published - 1 Jan 2016|
- Data mining
- Heart disease prediction
- K-nearest neighbor
- Machine to machine