Data Science and Intelligent Computing Techniques
Optimal Classification Process using Fuzzy C-Means Neural Network for Effective prediction of Cardiac arrest due to Diabetes
Authors: Prasadgouda Patil, Vijay Bhaskar Reddy and Ashokumar P
Publishing Date: 10-01-2023
ISBN: 978-81-955020-2-8
Abstract
In recent scores, diabetes mellitus (DM)is regarded as a chronic illness and one of the leading critical health challenges throughout the earth. About eighty percent of demise occurs because of DM(Type II) which could be avoided by the earlier diagnosis of persons with this threat. Nevertheless, presently machine learning techniques can be employed for diabetics’ detection very precisely. We are proffering a health care monitoring system comprising ECG sensors. The criteria that have a considerable volume of significance will be sensed by the ECG sensors that remain important for remote monitoring of the sick person. A mobile app observance will be employed for consistently monitoring the sick person’s ECG and diverse data extraction approaches will be executed upon the ECG wave for extracting features to properly prognosis heart illnesses. Hence, this study proffers the employment of a metaheuristic optimization algorithm called Real Coded Binary Ant Bee Colony (RC-BABC) for optimized feature choosing, and ReliefF methodology will be employed for excerpting the features and computing the features’ scores centered upon the disparities in feature values and class values betwixt nearby cases. An effectual attempt will be carried out for detecting cardiac demist at early phases emerging out of the intensity of DMin which feature prognosis before heart rate variability assessment will be executed. The DM’sfeatures would be analyzed out of the diabetic’s dataset for detecting the reason for abrupt cardiac arrest. Next, the excerpted features are classified employing the Fuzzy C-means Neural Network (FCNN). The performance analysis is carried out to exhibit that FCNN executes properly in prognosticating the illnesses. The proffered FCNN paradigm attains 97% and 84% of testing and training (t&t) accuracy, 93% and 82% of t&t specificity, 95% and 81% of t&t sensitivity and 92% and 85% of t&tF1-score.
Keywords
Keywords- cardiac arrest, diabetics, classification, neural network, preprocessing, optimization, feature score
Cite as
Prasadgouda Patil, Vijay Bhaskar Reddy and Ashokumar P, "Optimal Classification Process using Fuzzy C-Means Neural Network for Effective prediction of Cardiac arrest due to Diabetes", In: Satyasai Jagannath Nanda and Rajendra Prasad Yadav (eds), Data Science and Intelligent Computing Techniques, SCRS, India, 2023, pp. 1-12. https://doi.org/10.56155/978-81-955020-2-8-1