Supervised Weighting-Online Learning Algorithm for Short-Term Traffic Flow Prediction

Jeong, Young-Seon; Byon, Young-Ji; Castro-Neto, Manoel Mendonça de; Easa, Said M.

doi:10.1109/tits.2013.2267735

Cited by 257 publications

(113 citation statements)

References 37 publications

Supporting

Mentioning

112

Contrasting

Unclassified

Order By: Relevance

“…We compare the simulation results of the proposed scheme with the state-of-the-art prediction model, called online learning weighted support-vector regression (OLWSVR) which is proposed in [12] as the newest traffic estimation approach. Figure 6 illustrates the actual volume and estimated values of the proposed scheme and OLWSVR method and also shows the behaviour of the flow rate changing within current segment in three different charts.…”

Section: Simulation Resultsmentioning

confidence: 99%

“…In this case, a classification and regression trees (CART) model was proposed in [11] which predicts the short-term traffic volume at single locations in three steps: using the decision trees to classify the historical traffic states; founding the linear regression models and storing the weights in the leaf nodes of the trees model; and prediction the future traffic state through assigning the current state vector to the most congenial historical pattern and regression model. Also, the authors of [12] have proposed an OLWSVR approach for the short-term prediction of freeway traffic flow. The proposed scheme is an online weighted support-vector machine for regression, which combines an online supportvector machine for regression with a weighed learning method.…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Cluster-based traffic information generalization in vehicular ad-hoc networks

Arkian

Atani

Kamali

2014

7'th International Symposium on Telecommunications (IST'2014)

View full text Add to dashboard Cite

Vehicular Ad Hoc Network (VANET) is an emerging field of wireless networks that facilitates different applications such traffic information for participant vehicles and related authorities. However, deploying of such applications is mainly depending on the market penetration rate of this technology. In this paper, we propose a new 3-steps approach for estimation of traffic volume in a road segment based on actual volume of wireless-equipped vehicles. For this propose, we fist collect the traffic information for different groups of vehicles using a new clustering algorithm. Then, a chaining technique between the clusters transmits this information to the roadside cloud in the next step. Finally, we employ a machine learning method to generalization of the total traffic volume from the collected data. Performance of the proposed approach is evaluated using extensive simulation for different traffic densities, and the estimation accuracy of the proposed approach is shown through comparing to a state-of-the-art existing approach.

show abstract

Section: Simulation Resultsmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

Cluster-based traffic information generalization in vehicular ad-hoc networks

Arkian

Atani

Kamali

2014

7'th International Symposium on Telecommunications (IST'2014)

View full text Add to dashboard Cite

show abstract

“…First, the 20/80 confidence interval wraps closely the daily profile, indicating that the 12 series are very similar and that the daily profile is a representative summarization. Second, we observe that the daily profile shows two peaks, corresponding to the two rush hours: one in the morning (8-10 AM) corresponding to the daily commute towards work, and a second one in the afternoon (5)(6)(7)(8) corresponding to the end of the working day. Last, the daily profile allows to identify nonstandard days; for example, the blue line in Fig.…”

Section: B Daily Profile and Traffic Flow Mapmentioning

confidence: 89%

Motorway Traffic Flow Prediction using Advanced Deep Learning

Mihăiţă

et al. 2019

2019 IEEE Intelligent Transportation Systems Conference (ITSC)

View full text Add to dashboard Cite

Congestion prediction represents a major priority for traffic management centres around the world to ensure timely incident response handling. The increasing amounts of generated traffic data have been used to train machine learning predictors for traffic, however this is a challenging task due to inter-dependencies of traffic flow both in time and space. Recently, deep learning techniques have shown significant prediction improvements over traditional models, however open questions remain around their applicability, accuracy and parameter tuning. This paper proposes an advanced deep learning framework for simultaneously predicting the traffic flow on a large number of monitoring stations along a highly circulated motorway in Sydney, Australia, including exit and entry loop count stations, and over varying training and prediction time horizons. The spatial and temporal features extracted from the 36.34 million data points are used in various deep learning architectures that exploit their spatial structure (convolutional neuronal networks), their temporal dynamics (recurrent neuronal networks), or both through a hybrid spatiotemporal modelling (CNN-LSTM). We show that our deep learning models consistently outperform traditional methods, and we conduct a comparative analysis of the optimal time horizon of historical data required to predict traffic flow at different time points in the future.

show abstract

“…When the trained MARS model is applied on the testing data set, the generalization error might be quite large. In contrast, VS‐SVR predicts the response using SVR, which reduces the overfit effectively through introducing a ϵ ‐insensitivity loss function. In the other aspect, as the prediction module of VS‐SVR, SVR could yield encouraging results in a variety of experiments . However, SVR worked under low‐dimensional input in these literature.…”

Section: Experiments and Discussionmentioning

confidence: 99%

Urban traffic flow prediction: a spatio‐temporal variable selection‐based approach

Chen

Kong

et al. 2015

J of Advced Transportation

View full text Add to dashboard Cite

Summary Short‐term traffic flow prediction in urban area remains a difficult yet important problem in intelligent transportation systems. Current spatio‐temporal‐based urban traffic flow prediction techniques trend aims to discover the relationship between adjacent upstream and downstream road segments using specific models, while in this paper, we advocate to exploit the spatial and temporal information from all available road segments in a partial road network. However, the available traffic states can be high dimensional for high‐density road networks. Therefore, we propose a spatio‐temporal variable selection‐based support vector regression (VS‐SVR) model fed with the high‐dimensional traffic data collected from all available road segments. Our prediction model can be presented as a two‐stage framework. In the first stage, we employ the multivariate adaptive regression splines model to select a set of predictors most related to the target one from the high‐dimensional spatio‐temporal variables, and different weights are assigned to the selected predictors. In the second stage, the kernel learning method, support vector regression, is trained on the weighted variables. The experimental results on the real‐world traffic volume collected from a sub‐area of Shanghai, China, demonstrate that the proposed spatio‐temporal VS‐SVR model outperforms the state‐of‐the‐art. Copyright © 2015 John Wiley & Sons, Ltd.

show abstract

Supervised Weighting-Online Learning Algorithm for Short-Term Traffic Flow Prediction

Cited by 257 publications

References 37 publications

Cluster-based traffic information generalization in vehicular ad-hoc networks

Cluster-based traffic information generalization in vehicular ad-hoc networks

Motorway Traffic Flow Prediction using Advanced Deep Learning

Urban traffic flow prediction: a spatio‐temporal variable selection‐based approach

Contact Info

Product

Resources

About