Affiliation:
1. Karpagam Academy of Higher Education, Coimbatore, India
2. Galgotias University, India
Abstract
Patients can control, share, and manage their health records with family, friends, and healthcare professionals utilizing electronic health records (EHRs), which use an open channel, or the Internet. When a lot of data is available, DL methods show promise in these health applications. A distributed blockchain-based IoT system would benefit greatly from these ideas. This research proposes novel technique in Healthcare Data Based Feature Selection and Classification Using Blockchain and Machine Learning Architectures. The network has been secured using centralized blockchain sensor network. Here the input sensor-based healthcare data has been collected and processed for noise removal and smoothening. Then the processed data feature has been selected using Greedy Mixed Forward Colony Optimization feature selection. The suggested framework's superiority is supported by security research and experimental findings using the IoT-Botnet and ToN-IoT datasets. Proposed technique attained acc of 95%, precision of 85%, recall of 76%, F1 score of 63%, sec rate of 95%, DT Rate of 85%.