Cardio Disease Prediction using Deep Spectral Logistic Decision Neural Network

Nowadays, healthcare is growing rapidly due to the significant progress made in new technologies, such as the Internet of Things (IoT) and wearable devices. These devices are widely used to ensure remote patient monitoring. Smart devices offer innovative and enhanced features in smart healthcare systems and, thus, store a large amount of patients' sensitive information. Initially, we collect data from smart health devices deployed with IoT sensors such as pressure, pulse, heart rate etc. By utilizing these devices, valuable health data is obtained, and can now be filtered, analysed, and stored in electronic health records (EHRs). Initially we use Data Augmentation Normalization Filtering (DANF) method to eliminate the null values and inconsistency values. The Attribute Scaling Effect Rate (ASER) method is used to identify the marginal values of the collected features. Then Sensitive Correlation Feature Selection (SCFS) approach is to identify the importance of sensitive term weight. The DSLDNN approach is used to classify sensitive and non-sensitive disease prediction terms based on selected features. Finally, we evaluate the performance of the various techniques in the cardiac dataset. The results of the experiments indicate that the proposed method outperforms other previous techniques.