SAE Based DL Approach for ETF System

High rate of urbanization and economic growth has increased the volume of traffic in all part of the world. Intelligent traffic maintenance can, not only increase the movability of the vehicle but will also reduce the fuel cost. In summary, the research developed a new deep learning-based method for predicting traffic flow based on a stacked autoencoder model that takes into accounts the temporal dependencies in the dataset, and improves the representational capacity of the data. This improved representation will result in better forecasting accuracy of the model. This method uses the unsupervised learning methodology autoencoders to model more accurate prediction methodology for traffic prediction by stacking the autoencoders. The findings of this work demonstrate that this strategy performs better than conventional methods.