Enhanced tele ECG system using Hadoop framework to deal with big data processing

Abstract

Indonesia has high mortality caused by cardiovascular diseases. To minimize the mortality, we build a tele-ecg system for heart diseases early detection and monitoring. In this research, the tele-ecg system was enhanced using Hadoop framework, in order to deal with big data processing. The system was build on cluster computer with 4 nodes. The server is able to handle 60 requests at the same time. The system can classify the ecg data using decision tree and random forest. The accuracy is 97.14% and 98,92% for decision tree and random forest respectively. Training process in random forest is faster than in decision tree, while testing process in decision tree is faster than in random forest.