An Early Risk Prediction against Covid -19 Based On Adaptive Surf Scale Feature Selection and Sigmoid Recurrent Neural Network for Premature Precaution before Covid Second Wave

ABSTRACT

The development of new technologies, human nature, and food activities are completely against with nature in recent days. Due to this uncertainty, the new pandemic covid-19 has outbreak globally and spoiledthe human life. In this situation, the disease also has improved as a new modified version of influenza like Covid 19 (2.0). So, the Analysis of features becomes mandatory to identify the diseases based on big data analysis. Similarly, most people suffering from diabetes could not have a correct prediction to take proper treatments. Many health care suggestions and treatment handling methods do not predict the right information at the correct time to make a diagnosis. There is no premature process for the treatment based on non-predicted results. Hence, an Adaptive Surf Scale Feature Selection (SSFS) and Sigmoid Recurrent Neural Network Classification (SRNN) is proposed for improving the early risk prediction against the covid. Initially, preprocessing is carried out to verify the records. Then, future selection is carried out using the surf scale weightage factor, which analyzes the Covid Influence Rate (CIR) and the marginal weight is identified. Each identified weight is ruled into Inter Class Sigmoid Function (ICSF) to activate the iteration recurrent neural network. Finally, the classifier trains the features into a feed-forward layer to produce the classified result. The proposed system produces high performance compared to the previous system as well as recommendations for premature diagnoses.Copyright © 2020 Ubiquity Press. All rights reserved.