A Novel Optimized LSTM Networks for Traffic Prediction in VANET
Unknown authors
Abstract
A network of vehicular cyber-physical systems uses wireless communications or Internet for efficient data transfer, which includes, safety, transportation details, mobility and sustainability.With the advent of vehicular IoT, smart transportation systems play a vital role in today's life for the prediction of traffic flows and efficient data transfer.With the advent of machine learning algorithms, prediction of traffic flow has reached its new dimension but still usage of single model machine learning algorithms needs improvisation in terms of prediction accuracy.Hence this paper proposes the new model of predicting the traffic flows based on the hybrid optimized learning algorithm, which integrates the BAT optimized and LSTM algorithm (BAT-LSTM).First, BAT Algorithm is applied to obtain the hyper parameters of each LSTM predictor.Again, LSTM prediction is trained using the training samples obtained by the optimized BAT algorithm.More than 100 hours of real-time traffic datasets were analysed and used for evaluating the proposed hybrid algorithms, which was then experimented on SUMO with OMNET++ platforms.The empirical study demonstrates that the proposed approach outperformed other existing approaches in regards of accuracy, sensitivity, and selectivity.It infers that the proposed approach is extraordinary performance for traffic prediction and management systems.