Short-term speed predictions exploiting big data on large urban road networks

Abstract

Big data from floating cars supply a frequent, ubiquitous sampling of traffic conditions on the road network and provide great opportunities for enhanced short-term traffic predictions based on real-time information on the whole network. Two network-based machine learning models, a Bayesian network and a neural network, are formulated with a double star framework that reflects time and space correlation among traffic variables and because of its modular structure is suitable for an automatic implementation on large road networks. Among different mono-dimensional time-series models, a seasonal autoregressive moving average model (SARMA) is selected for comparison. The time-series model is also used in a hybrid modeling framework to provide the Bayesian network with an a priori estimation of the predicted speed, which is then corrected exploiting the information collected on other links. A large floating car data set on a sub-area of the road network of Rome is used for validation. To account for the variable accuracy of the speed estimated from floating car data, a new error indicator is introduced that relates accuracy of prediction to accuracy of measure. Validation results highlighted that the spatial architecture of the Bayesian network is advantageous in standard conditions, where a priori knowledge is more significant, while mono-dimensional time series revealed to be more valuable in the few cases of non-recurrent congestion conditions observed in the data set. The results obtained suggested introducing a supervisor framework that selects the most suitable prediction depending on the detected traffic regimes

Improving Coverage Rate for Urban Link Travel Time Prediction Using Probe Data in the Low Penetration Rate Environment

“…Data based models can be further grouped into parametric models (e.g., linear regression, Bayesian net, and time-series models) and non-parametric models (e.g., neural networks (NN), pattern recognition and support vector machine models). Under complicated traffic situations such as urban networks, non-parametric models outperform parametric models and it is relatively easy for non-parametric models to be extended from one application to another [2,21,22]. However, the influence of penetration rate is more significant for non-parametric models, such as the kNN-based model [9,21,23] and the PF-based model [7,24], because these models require time-sequential samples that cover exactly the same time interval and have exactly the same sample length.…”

A Hybrid Approach Based on Variational Mode Decomposition for Analyzing and Predicting Urban Travel Speed

“…ey generally have reported that a machine learning-based approach, such as arti cial neural network (ANN) [7], the support vector machine (SVM) [8,9], and the k-nearest neighbor method (KNN) [10][11][12], outperform the parametric statistical approach due to their ability to identify the nonlinear e ects through exible restructuring [13]. However, the existing learning-based approaches have di culty dealing with the uncertainty of urban tra c conditions, and the di culty is compounded in unstable congestion [14]. Another characteristic of urban travel speed is spatiotemporal heterogeneity.…”

“…However, since the influences of links on their neighbors change with time and they are heterogeneous in space, only generally well-performed spatiotemporal variables (i.e., the number and location of links and the optimal timelag) would be selected at a given point in time. ese limitations cause poor performance, particularly in nonrecurrent conditions due to irrelevant variables [14,16].…”

Insights from self-organizing maps for predicting accessibility demand for healthcare infrastructure

“…A common approach in these studies is to derive statistical relationships between the basic characteristics of catchment populations. However, whilst valuable insights can be drawn from such analysis, basic descriptors such as median or grouped income do not always provide sufficient nuance for in-depth analysis of accessibility changes across populations.The increasing prevalence of 'big data' in the urban sphere makes it possible to observe, analyse and predict human behaviour at increasingly fine scales [27][28][29]. Advances in computational data mining based on machine learning (ML) techniques are also enabling researchers to make better sense of these rich datasets [30].…”

“…SOM algorithms have been applied to a wide range of disciplines (see reviews in Reference [33]), from analysis of volcano seismic spectra [34] and atmospheric aerosol tracking [35] to land value prediction [36]. Other ML techniques (e.g., neural networks, random forest classifiers) have also been applied to urban transportation problems, including for predicting traffic flows [27] and travel mode choice [37] and for clustering transit card usage [38]. However, barring some exceptions (e.g., [39][40][41][42]), SOM analysis has had a relatively limited impact in the social sciences, despite the possibilities it offers to intuitively visualise complex socio-demographic patterns, which can notably be helpful in the context of policy-making [42,43].In this study, we aim to relate socio-economic factors to accessibility across a city experiencing rapid population growth.…”