Air Traffic Data Analysis Using Recurrent Neural Network (RNN) Classifier During COVID-19

Air navigation achieves a worldwide transportation system for economic growth, creates jobs, and facilitates international traffic and sightseeing. In recent times, COVID-19 has been considered as a most dangerous disease that affects many people. When people travel on air flights, COVID-affected people can easily spread fluids into the air. Thus, prediction of air traffic data is considered a significant problem in identifying COVID disease. Few research works have been developed to perform traffic data analysis but, the accuracy of data analysis was not improved with less time. The Concept Drift Regressive Recurrent Neural Network classifier (CDR-RNNC) model is proposed to analyze air traffic data during COVID-19. The main aim of the developed CDR-RNNC model is to identify air traffic with less time and higher accuracy to perform accurate traffic data prediction. Initially, air traffic data with different features are taken as input and fed into first hidden layer for presenting feature identification to attain traffic data analysis with higher accuracy. Based on identified data features, air traffic data classification is performed by applying Deming regression analysis and identifying the relationship between traffic data. Then, analyzed air traffic data is used to estimate concept drift for classifying traffic data as lesser or higher. Finally classified results for predicting the spread of COVID-19 disease with less time. Experimental evaluation of the CDR-RNNC model is conducted based on metrics such as air traffic data prediction accuracy, time complexity, and false alarm rate with respect to number of air traffic data. The experiment results demonstrate that the developed model increases prediction accuracy with minimum amount of time required for air traffic data prediction when compared to state-of-the-art works.