Survey of Traffic Prediction by LSTM Using Machine Learning Techniques

Abstract

Machine learning technologies are fast-growing domain for prediction. The main reason for traffic is traffic signals, accident, weather and road repair.In Real-time mostly traffic data are generated exponentially and we have to enhance data transportation using big data concepts. This fact encouraged as to build a better traffic flow prediction model. First, we should collect the large number of traffic prediction journals and study their work pattern or algorithm. In this work, we are planned to use machine learning and deep learning algorithm and LSTM algorithm.We are proposed artificial recurrent neural network (LSTM) for traffic prediction. We gathered large amount of data for analyzing traffic flow and based on that we compare best data flow then finally we got an expected prediction result.

We survived several authors’ works and demonstrated our results in section III using python simulation. We improved 6-7% percentage throughput and minimized interference also.