Application of LSTM in Short-term Traffic Flow Prediction

Abstract

As urbanization intensifies, the status of the traffic situation predict is becoming more and more prominent. The urban traffic flow is influenced by many factors and is characterized by strong randomness. This paper combines MSE and Adam to construct a linear LSTM to realize the prediction of short-term traffic flow based on time series. The experiment result shows that LSTM can gain the periodic features of the traffic flow. It has small error and high precision for the short-term prediction of the traffic flow based on time series, which verifies the validity of LSTM.