Missing Data Analysis in Traffic Monitoring System Using LSTM

Abstract

A dependable traffic route optimization system and an accurate traffic simulation model is crucial for effective traffic management system. The creation of a more efficient traffic management system is hampered by the absence of a realistic traffic simulation data. Missing data is one of the fundamental causes of hindrance, because it will certainly result in inaccurate predictions of congestion levels and less effective reconfiguration. Both future traffic data mining and real-time traffic monitoring are impacted by these missing numbers. In this paper we used transfer learning-based algorithm to fill in the gaps of those missing data. This process of data recreation will help the existing traffic management system by improving its accuracy and efficiency which will directly improve the working of traffic simulation models and traffic route optimization systems.