Sensor-Based Technique for The Identification of Cardiac Disorders Utilising Feature Extraction and Artificial Neural Networks

By combining techniques for signal processing and feature extraction, this study suggests a novel approach to person identification. It uses an artificial neural network (ANN) and 10 metal oxide semiconductor sensors to identify each person's distinct scent patterns. The first step in using ANN patterns is to scan and acquire sensor data. You must first scan and collect sensory information from the sensor data before using ANN patterns to create patterns from it. Each participant in the several studies, which cover a range of periods and include 5, 10, 15, and 20 people, is identified and scanned for more than a thousand unique traits. Arduino receives analogue sensor inputs and uses changing periods to convert them to digital form. The architecture has to be trained using the recently generated dataset. Metrics like sensitivity, f-measures, specificity, and accuracy are used to evaluate the proposed model for human odor identification. Research uses assessment metrics, and the results show that this model is generally accurate to within 15% of the true value. The findings point to potential applications of feature extraction techniques to improve overall person identification and human odor detection.