Accurate and Interpretable Bayesian MARS for Traffic Flow Prediction

Abstract

Current research on traffic flow prediction mainly concentrates on generating accurate prediction results based on intelligent or combined algorithms but ignores the interpretability of the prediction model. In practice, however, the interpretability of the model is equally important for traffic managers to realize which road segment in the road network will affect the future traffic state of the target segment in a specific time interval and when such an influence is expected to happen. In this paper, an interpretable and adaptable spatiotemporal Bayesian multivariate adaptive-regression splines (ST-BMARS) model is developed to predict short-term freeway traffic flow accurately. The parameters in the model are estimated in the way of Bayesian inference, and the optimal models are obtained using a Markov chain Monte Carlo (MCMC) simulation. In order to investigate the spatial relationship of the freeway traffic flow, all of the road segments on the freeway are taken into account for the traffic prediction of the target road segment. In our experiments, actual traffic data collected from a series of observation stations along freeway Interstate 205 in Portland, OR, USA, are used to evaluate the performance of the model. Experimental results indicate that the proposed interpretable ST-BMARS model is robust and can generate superior prediction accuracy in contrast with the temporal MARS model, the parametric model autoregressive integrated moving averaging (ARIMA), the state-of-the-art seasonal ARIMA model, and the kernel method support vector regression. Index Terms-Bayesian inference, interpretable model, Markov chain Monte Carlo (MCMC), multivariate adaptive-regression splines (MARS), spatiotemporal relationship analysis, traffic flow prediction.

Bag of Contextual-Visual Words for Road Scene Object Detection From Mobile Laser Scanning Data

“…p H i = (x i , y i ) and p H j = (x j , y j ) are the coordinates of the centroids on the XY plane; p V i = z i and p V j = z j are the z coordinates of the centroids; I n i and I n j are the interpolated normalized intensities of the points in voxels i and j, respectively. The interpolated normalized intensity of a voxel can be computed using (8). σ 2 H , σ 2 V , and σ 2 I are the variances of the horizontal, vertical, and intensity distributions, respectively.…”

“…and safety warning systems [1], [2], autonomous driving [3]- [5], and traffic flow monitoring and prediction [6]- [8]. In addition, accurate, real-time information regarding current road conditions, traffic flow, and the surrounding environment is of great significance and necessity to the Intelligent Transportation Systems.…”

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

“…To verify the prediction ability of the proposed VS‐SVR model, the following five traffic prediction models were implemented and evaluated in the experiments: AR: AR is widely applied in traffic flow prediction as a time series modeling tool. In most studies, AR served as a simple baseline to appraise the prediction ability of the new models . In this paper, the three‐order AR was built to predict the testing set using the temporal traffic states. MARS: Despite that MARS is used as a variable selection technique in this paper, it can be employed to predict the short‐term traffic flow .…”

Spatiotemporal variable and parameter selection using sparse hybrid genetic algorithm for traffic flow forecasting

“…To compare the prediction performance of various forecasting methods, two widely used criteria, namely, RMSE and mean absolute percentage error (MAPE), 17 are adopted in this study. In addition, since Lasso and our SHGA-LSSVR both can select spatiotemporal variables, the number of selected variables is also recorded which indicates the prediction performance can be achieved with typically much fewer variables.…”