Open Access
Traffic speed prediction using big data enabled deep learning
Shuo Wang
- 01 Jan 2018
4
TL;DR: The proposed long-term and short-term traffic speed prediction models can be combined as a multilayer decision supporting system to provide traffic management an opportunity to operate proactively.
read more
Abstract: The objective of the proposed study is to predict traffic speeds at a route level so that the traffic management has a chance to operate proactively. A distributed file system and parallel computing platform is used to store the big data sets of statewide traffic and weather data in a fault-tolerant way and process the big data in a timely manner. Traffic speed prediction problem is studied at two levels, and two deep networks are proposed accordingly: a fully convolutional deep network for long-term speed prediction and a hybrid long short-term memory (LSTM) network for short-term speed prediction. The fully convolutional deep network utilizes both weather information and historical traffic speeds to make long-term traffic speed predictions, and a trained model can be transferred to predict traffic speed at any spatial-temporal scale. The hybrid LSTM network utilizes the previous traffic speeds on the current day as well as historical traffic speeds to make short-term speed predictions, and a trained model can be used to predict speeds at any timestamps ahead in a streaming fashion. The proposed long-term and short-term traffic speed prediction models can be combined as a multilayer decision supporting system to provide traffic management an opportunity to operate proactively.
read more
Chat with Paper
AI Agents for this Paper
Find similar papers on Google Scholar, PubMed and Arxiv
Write a critical review of this paper
Analyze citations of this paper to find unaddressed research gaps
Citations
Predicting travel times for the south jersey real-time motorist information system
Steven Chien,Xiaobo Liu,Kaan Ozbay +2 more
- 01 Jan 2003
TL;DR: In this paper, the authors developed a dynamic travel time prediction model for the South Jersey Motorist Real-time Information System (SJMIRIS) by integrating traffic flow theory, traffic simulation, measurement and application of collected data.
1
Patent
PCM A Real-Time Detection System And Method On Flush Reload Attack Using PCM
Cho Jonghyeon,Kim Taehyeon,Shin Youngjoo +2 more
- 01 Jun 2020
TL;DR: In this article, a system for detecting a flush+reload cache subchannel attack in real time using a performance counter attack (PCM) and a method thereof is presented.
A Deep Learning-Based Utilization Improvement Framework for Abdullah Hashim Industrial Gases & Equipment Co. Ltd Transportation Network
TL;DR: In this article , a deep learning-based model was developed to improve the efficiency of the constrained transportation network of an industrial gas company in conjunction with historical data using an example case of the gas industry in Jeddah, Saudi Arabia.
References
Long short-term memory
TL;DR: A novel, efficient, gradient based method called long short-term memory (LSTM) is introduced, which can learn to bridge minimal time lags in excess of 1000 discrete-time steps by enforcing constant error flow through constant error carousels within special units.
99K
Going deeper with convolutions
Christian Szegedy,Wei Liu,Yangqing Jia,Pierre Sermanet,Scott Reed,Dragomir Anguelov,Dumitru Erhan,Vincent Vanhoucke,Andrew Rabinovich +8 more
- 07 Jun 2015
TL;DR: Inception as mentioned in this paper is a deep convolutional neural network architecture that achieves the new state of the art for classification and detection in the ImageNet Large-Scale Visual Recognition Challenge 2014 (ILSVRC14).
Reducing the Dimensionality of Data with Neural Networks
TL;DR: In this article, an effective way of initializing the weights that allows deep autoencoder networks to learn low-dimensional codes that work much better than principal components analysis as a tool to reduce the dimensionality of data is described.
A fast learning algorithm for deep belief nets
TL;DR: A fast, greedy algorithm is derived that can learn deep, directed belief networks one layer at a time, provided the top two layers form an undirected associative memory.
Traffic Flow Prediction With Big Data: A Deep Learning Approach
TL;DR: A novel deep-learning-based traffic flow prediction method is proposed, which considers the spatial and temporal correlations inherently and is applied for the first time that a deep architecture model is applied using autoencoders as building blocks to represent traffic flow features for prediction.
3.1K