SECURED PREDICTION MODEL USING IOT BASED SMART DEVICES FOR HEALTHCARE ASSISTANCE

Abstract

Over the past three decades, the internet's expansion and the advancement of technology have altered the world. The Internet of Things (IoT) is a network of networked devices that can collect and share data using built-in sensors and communication protocols. There are many different device types included in the IoT, some of which could be useful for scientific research. The Internet of Things (IoT) enhances the Web by deploying ubiquitous devices with embedded identification, sensing, and data exchange capabilities. These intelligent objects provide the basis of adaptable cyber-physical networks that provide a platform for networked data transmission. In this work, we suggest a novel method for predicting and analyzing stroke severity in older adults over 65 that is based on the National Institutes of Health Stroke Scale (NIHSS). The input dataset is raw and may contain missing values and redundant packets when first gathered. After data has been acquired, normalize the data using Label Encoder Min-Max. Advanced Encryption Standard (AES) is a popular symmetric encryption technique that offers secure data encryption and decryption after test and training. With the use of the Linear Discriminant Analysis (LDA) method, we were able to extract features from the data and acquire different data properties. Next, we used Singular Value Decomposition to choose the features. Finally, we suggested that Genetic Decision Classifier (GDC) be used to forecast an individual's health condition based on the features chosen. Our method may produce 96% accuracy, 93% precision, 97% recall, and 95% f1-score.