Fuzzy Inference with Enhanced Convolutional Neural Network Based Classification Framework for Predicting Heart Attack Using Sensor Data

Heart attack is becoming a common life-threatening disease in the current lifestyle of the people. There is a lack of an automatic mechanism to detect it using medical sensor data. Because processing medical sensor data is tedious and leads to more processing overhead. Hence, we are proposing a framework to process the data by mining the important patterns in it i.e., we designed a framework to process the data sensed by various sensors that can be deployed on the human body in the form of gadgets using the technologies like Internet of Things and Machine learning. This paper presents a Fuzzy Inference with a Modified Convolutional Neural Network framework for heart attack prediction. We trained and tested our designed framework on the medical sensor data and achieved good prediction accuracy. This framework offers a foundation for developing a system of decision support that has the potential for learning and the ability to cope with disease management vagueness and unstructuredness. The result is compared with a few existing methods like SVM, K-Nearest Neighbor, Logistic Regression, and Convolutional Neural Network to show improved classification accuracy.