{"title":"Towards the Design of Smart Vehicular Traffic Flow Prediction","authors":"A. Boukerche, Jiahao Wang","doi":"10.1145/3479241.3486701","DOIUrl":null,"url":null,"abstract":"Thanks to the fast development of computing hardware and Machine Learning-based (ML) model, many impressive prediction models have been proposed under the topic of traffic flow prediction. While ML models highly improve the accuracy of the prediction system, it has higher time consumption on the training phase when being applied to a large traffic network, compared to traditional time-series models. The other thing we should consider when predicting the traffic flow in a large traffic network is to utilize the spatial correlation among the detectors. To solve above problems, we will provide a traffic flow prediction solution in this paper. The solution has three parts: a hybrid prediction model based on Graph Convolutional Network (GCN) and Recurrent Neural Network (RNN), which can extract spatial-temporal features from dataset; a prediction strategy for multi-step prediction; an efficient training strategy for prediction on large-scale network.","PeriodicalId":349943,"journal":{"name":"Proceedings of the 19th ACM International Symposium on Mobility Management and Wireless Access","volume":"111 1","pages":"0"},"PeriodicalIF":0.0000,"publicationDate":"2021-11-22","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":"4","resultStr":null,"platform":"Semanticscholar","paperid":null,"PeriodicalName":"Proceedings of the 19th ACM International Symposium on Mobility Management and Wireless Access","FirstCategoryId":"1085","ListUrlMain":"https://doi.org/10.1145/3479241.3486701","RegionNum":0,"RegionCategory":null,"ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"","JCRName":"","Score":null,"Total":0}
引用次数: 4
Abstract
Thanks to the fast development of computing hardware and Machine Learning-based (ML) model, many impressive prediction models have been proposed under the topic of traffic flow prediction. While ML models highly improve the accuracy of the prediction system, it has higher time consumption on the training phase when being applied to a large traffic network, compared to traditional time-series models. The other thing we should consider when predicting the traffic flow in a large traffic network is to utilize the spatial correlation among the detectors. To solve above problems, we will provide a traffic flow prediction solution in this paper. The solution has three parts: a hybrid prediction model based on Graph Convolutional Network (GCN) and Recurrent Neural Network (RNN), which can extract spatial-temporal features from dataset; a prediction strategy for multi-step prediction; an efficient training strategy for prediction on large-scale network.