Citywide Traffic Flow Prediction Based on Multiple Gated Spatio-temporal Convolutional Neural Networks

Abstract

Traffic flow prediction is crucial for public safety and traffic management, and remains a big challenge because of many complicated factors, e.g., multiple spatio-temporal dependencies, holidays, and weather. Some work leveraged 2D convolutional neural networks (CNNs) and long short-term memory networks (LSTMs) to explore spatial relations and temporal relations, respectively, which outperformed the classical approaches. However, it is hard for these work to model spatio-temporal relations jointly. To tackle this, some studies utilized LSTMs to connect high-level layers of CNNs, but left the spatio-temporal correlations not fully exploited in low-level layers. In this work, we propose novel spatio-temporal CNNs to extract spatio-temporal features simultaneously from low-level to high-level layers, and propose a novel gated scheme to control the spatio-temporal features that should be propagated through the hierarchy of layers. Based on these, we propose an end-to-end framework, multiple gated spatio-temporal CNNs (MGSTC), for citywide traffic flow prediction. MGSTC can explore multiple spatio-temporal dependencies through multiple gated spatio-temporal CNN branches, and combine the spatio-temporal features with external factors dynamically. Extensive experiments on two real traffic datasets demonstrates that MGSTC outperforms other state-of-the-art baselines.

Modeling Dynamic Spatio-Temporal Correlations for Urban Traffic Flows Prediction

“…Most of the research scholars applied CNN for different kinds of problems specifically in the computer vision area [9]. According to [20] presented residual learning that allows those networks which have a [13] explicit explicit explicit DHSTNet [5] explicit explicit explicit STD-Net explicit explicit explicit more deep structure. According to [21], they used a recurrent neural network (RNN) for sequence learning tasks as well as researchers used long short term memory (LSTM) that enables CNN to understand long term temporal dependency.…”

Task Allocation on Layered Multiagent Systems: When Evolutionary Many-Objective Optimization Meets Deep Q-Learning

“…Fortunately, the nowadays deep learning algorithms are well known for their capabilities of handling complicated situations, and therefore are well suited to be applied to deal with allocation issues for MASs. Some upto-date deep learning methods include, but are not limited to, Deep Q-Network (DQN) [42], Convolutional Neural Network [9], Feedforward Neural Network [56] and Recurrent Neural Network [1], which are ideal candidate algorithms for further improving the efficiency of task allocation and execution [33]. These networks in deep-learning process usually combined with back-propagation [22], and Stochastic Gradient Descent (SGD) [3] methods.…”

Advantages of Combining Factorization Machine with Elman Neural Network for Volatility Forecasting of Stock Market

“…Artificial neural network has the advantages on learning from sample data and capturing the nonlinear relations among interconnected neurons through training mode [3]. It is capable of dealing with nonlinear highdimensional data and approximating any nonlinear functions with arbitrary precision [4][5][6][7]. Particularly, the simple recurrent network, i.e., Elman neural network (Elman NN) [8] has shown its stronger ability as it has the characteristic of time-varying.…”