Implementing A Deep Learning Framework for Short Term Traffic Flow Prediction

Traffic congestion in highway systems has become more serious. Recently, with the development of advanced technologies for intelligent transportation system, many studies focus on proposing new models and algorithms for time series analysis to predict short-term traffic flow. In this paper, we take a comprehensive study in implementing a deep learning framework for the short term traffic flow prediction in highway systems. Specifically, we apply long-short term memory for analyzing data in Gyeongbu Expressway of Korean transportation system. Experiments indicate promising results for predicting short term traffic flow in highway systems.