Deep and Embedded Learning Approach for Traffic Flow Prediction in Urban Informatics

Zheng, Zibin; Yang, Yatao; Liu, Jiahao; Dai, Hong-Ning; Zhang, Yan

doi:10.1109/tits.2019.2909904

Cited by 128 publications

(46 citation statements)

References 40 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Furthermore, ref. [24] shows that the composite models present the excellent performance than the sole DL models. Therefore, we summarize the existing approaches having two major limitations.…”

Section: Input Featurementioning

confidence: 92%

Wide-Attention and Deep-Composite Model for Traffic Flow Prediction in Transportation Cyber–Physical Systems

Zhou

Dai

Wang

2021

IEEE Trans. Ind. Inf.

Self Cite

View full text Add to dashboard Cite

Recently, traffic flow prediction has drawn significant attention because it is a prerequisite in intelligent transportation management in urban informatics. The massively-available traffic data collected from various sensors in Transportation Cyber-Physical Systems brings the opportunities in accurately forecasting traffic trend. Recent advances in deep learning shows the effectiveness on traffic flow prediction though most of them only demonstrate the superior performance on traffic data from a single type of vehicular carriers (e.g., cars) and does not perform well in other types of vehicles. To fill this gap, we propose a wide-attention and deep-composite (WADC) model consisting of a wide-attention module and a deep-composite module in this paper. In particular, the wide-attention module can extract global key features from traffic flows via a linear model with selfattention mechanism. The deep-composite module can generalize local key features via Convolutional Neural Network component and Long Short-Term Memory Network component. We also perform extensive experiments on different types of traffic flow datasets to investigate the performance of WADC model. Our experimental results exhibit that WADC model outperforms other existing approaches.

show abstract

Section: Input Featurementioning

confidence: 92%

Wide-Attention and Deep-Composite Model for Traffic Flow Prediction in Transportation Cyber–Physical Systems

Zhou

Dai

Wang

2021

IEEE Trans. Ind. Inf.

Self Cite

View full text Add to dashboard Cite

show abstract

“…Meanwhile, the computational complexity of one GRU is

scriptO ()(, M b_{m} d_{h}^{2})

, where b m is the length of the mini‐batch sequence and d h is the hidden state size, respectively [27]. The computational complexity of the CNN and GRU joint model can be estimated by summarising up the computational costs of CNN and GRU components, respectively [28]. Among them, one can see that CNN dominates the overall computational complexity.…”

Section: Deep Learning Model For Wind Speed Forecasting At Turbine mentioning

confidence: 99%

Deep learning approach for wind speed forecasts at turbine locations in a wind farm

Kou

Wang

Liang

et al. 2020

IET Renewable Power Generation

View full text Add to dashboard Cite

In a wind farm, individual turbines disturb the wind field by generating wakes, so wind speeds at various turbine locations are different. From the perspective of wind farm control, there is an interest in dynamic optimization of the power reference for each individual wind turbine, and the wind speed forecast at each turbine location is hence required. This paper develops a joint model of convolutional neural network (CNN) and the gated recurrent units (GRU) to forecast the wind speed at turbine locations. This model employs a two-layer architecture. At the lower-layer, the spatial features are automatically extracted by CNN. The extracted spatial features describe the spatial correlations among multiple wind turbines. At the upperlayer, GRU learns the temporal correlations across the extracted spatial features. This joint model is trained in an integrated manner. A salient characteristic of this model is that it extracts high-level spatial-temporal features from wind data. These automatically learnt features capture the spatial-temporal wind dynamics and interactions in a wind farm, thus being informative and appropriate for the forecasting at specific turbine locations. The simulation on actual data demonstrates the effectiveness of the presented model.

show abstract

“…In contrast to conventional ML approaches, deep learning (DL) approaches have the merits in capturing the complicated features from massive data [15]. For example, DL approaches have been used in sentimental analysis [16], electricity-theft detection [17] and traffic flow prediction [18].…”

Section: B Deep Learning Approachesmentioning

confidence: 99%

Precious Metal Price Prediction Based on Deep Regularization Self-Attention Regression

Zhou

Song

et al. 2020

IEEE Access

Self Cite

View full text Add to dashboard Cite

It is non-trivial to predict the prices of precious metals since a number of factors can affect the fluctuations of precious metal prices. Either parametric models or machine learning models cannot accurately forecast the precious metal prices. Though deep learning approaches show their strengths in extracting key features from complicated data, they have the limitations of learning localization and losing some temporal and spatial features. The recent advances in attention mechanisms bring the opportunities to overcome the limitation of deep learning models. In this paper, we originally propose a Regularization Self-Attention Regression Model for precious metal price prediction. In particular, the proposed RSAR model consists of convolutional neural network (CNN) component and Long Short-Term Memory Neural Networks (LSTM) component. Integrating with self-attention mechanism, this model can extract both spatial and temporal features from precious metal price data. Meanwhile, the proper configuration of regularization functions can also lead to the further performance improvement. Extensive experiments on realistic precious metal price dataset show that our proposed approach outperforms other conventional machine learning and deep learning methods. INDEX TERMS Long short-term memory, convolutional neural network, attention mechanism, financial data analysis, deep learning.

show abstract

Deep and Embedded Learning Approach for Traffic Flow Prediction in Urban Informatics

Cited by 128 publications

References 40 publications

Wide-Attention and Deep-Composite Model for Traffic Flow Prediction in Transportation Cyber–Physical Systems

Wide-Attention and Deep-Composite Model for Traffic Flow Prediction in Transportation Cyber–Physical Systems

Deep learning approach for wind speed forecasts at turbine locations in a wind farm

Precious Metal Price Prediction Based on Deep Regularization Self-Attention Regression

Contact Info

Product

Resources

About