Design of COVID19 Detection based on Relative Eccentric Feature Selection using Deep Vectorized Regressive Neural Network for Corona Virus

• Published in: 2023 IEEE International Conference on Integrated Circuits and Communication Systems (ICICACS) pp. 1 - 6
• Main Authors: Mishra, Saket, Mantri, Archana
• Format: Conference Proceeding
• Language: English
• Published: IEEE 24.02.2023
• Subjects: COVID-19; COVID19; Deep learning; Deep Vectorized Regressive Neural Network (DVRNN; Design methodology; Feature extraction; Integrated circuits; Neural networks; Pandemics; preliminary process; Relative Eccentric Feature Selection (REFS)
• DOI: 10.1109/ICICACS57338.2023.10099777

The world has seen various diseases in different variants, numerous pandemics in the twentieth century like COVID-19. Fly infections are the fundamental driver of contaminations. COVID-19 declared a global pandemic with major impacts on economies and societies around the world. The diagnosis of COVID19 or non-COVID-19 cases early detection at the correct separation early stages of disease are one of the main concerns of the current coronavirus pandemic. At present, accurate detection of coronavirus disease usually takes a long time and is prone to human error. To address this problem, the proposed Deep learning and Design of COVID19 detection based on Relative Eccentric Feature Selection (REFS) Using Deep Vectorized Regressive Neural Network (DVRNN) for corona virus the early detection of the COVID19 virus. Initially collects the COVID19 sample test dataset, then the raw dataset trained into preliminary process is used to remove unwanted noise. After that preliminary processed dataset trained into the feature selection process is done to identify the best features of COVID19 using Ensemble recursive selection. Further, the proposed DVRNN algorithm is done to classify the accurate detection of coronavirus. The proposed model would be useful for the Timely and accurate identification of various stages of coronavirus. Therefore, it can detect the accurate results of COVID19 effectively and accomplish good performance compared with previous methods.