Short-term traffic prediction consists a crucial component in intelligent transportation systems. With the explosion of automated traffic monitoring sensors and the flourishing of deep learning techniques, a growing body of deep neural network models have been employed to tackle this problem. In particular, convolutional neural networks (CNN) and long short-term memory (LSTM) recurrent networks have demonstrated their advantages in modeling and predicting the spatiotemporal evolution of traffic flows. In this paper, we propose a novel Convolutional LSTM neural network architecture for multi-lane shortterm traffic prediction. Compared to existing methods, we highlight the importance of (1) applying multiple features to characterize traffic conditions; (2) explicitly considering the routing between neighbouring lanes and downstream/upstream traffics; and (3) predicting multiple time-step traffic in a rolling-prediction manner. Experiments on 10 months 5-minute interval observations of the US I-101 Northern freeway at California Bay Area verify the proposed model. The results show that our model has considerable advantages in predicting multi-lane short-term traffic flow. INDEX TERMS Short-term traffic flow prediction, multi-lane traffic flow, convolutional LSTM.

Boosted by the growing logistics industry and digital transformation, the sharing warehouse market is undergoing a rapid development. Both supply and demand sides in the warehouse rental business are faced with market perturbations brought by unprecedented peer competitions and information transparency. A key question faced by the participants is how to price warehouses in the open market. To understand the pricing mechanism, we built a real world warehouse dataset using data collected from the classified advertisements websites. Based on the dataset, we applied machine learning techniques to relate warehouse price with its relevant features, such as warehouse size, location and nearby real estate price. Four candidate models are used here: Linear Regression, Regression Tree, Random Forest Regression and Gradient Boosting Regression Trees. The case study in the Beijing area shows that warehouse rent is closely related to its location and land price. Models considering multiple factors have better skill in estimating warehouse rent, compared to singlefactor estimation. Additionally, tree models have better performance than the linear model, with the best model (Random Forest) achieving correlation coefficient of 0.57 in the test set. Deeper investigation of feature importance illustrates that distance from the city center plays the most important role in determining warehouse price in Beijing, followed by nearby real estate price and warehouse size.

The condition monitoring and fault detection of rolling bearing are of great significance to ensure the safe and reliable operation of rotating machinery system.In the past few years, deep neural network (DNN) has been recognized as an effective tool to detect rolling bearing faults. However, It is too complex to directly feed the original vibration signal to the DNN neural network, and the accuracy of fault identification is not high. By using the signal preprocessing technology, the original signal can be effectively removed and preprocessed without losing the key diagnosis information. In this paper, a new EEMD-LSTM bearing fault diagnosis method is proposed, which combines the signal preprocessing technology with the EEMD method that can get clear fault feature signals, and LSTM technology to extract fault features automatically that improves the efficiency of fault feature extraction. In the case of small sample size, this method can significantly improve the accuracy of fault diagnosis.

This study investigates environment sensitive and perishable products (ESPPs) logistics problem, which is called cold chain logistics problem (CCLs). Based on a comprehensive literature review, we found that there is much room to improve regarding of the risks management in cold chain logistics, that is, the development of a comprehensive cold chain logistics design methodology should considered uncertainty sources and risk exposures. In this study, we propose a neural network model to illustrate the problems. Firstly, the paper develops input indicators at different points in cold chain logistics to examine the effects of environment fluctuations including temperature control, humidity monitoring, the temperature interruption time and electric vehicle mapping, etc; secondly, the improved neural network algorithm can achieve model convergence, including the increase of momentum term, the adjustment of learning rate and the change of error function. At last, through simulation, this study shows that comprehensive risk prediction of cold chain logistics will be calculated based on the input indicators using the improved neural network algorithm, and the predictive value is accurate. So not only the analyzing of kinds of cold chain logistics indicators can be realized through the Neural Network model, but we can take priorities resorting to the predictive results accordingly.

The prediction of travel mode preference, like many other choice prediction problems, may depend on categorical features of the choice options or the choice makers. Such categorical features need to be meaningfully encoded for better modeling and understanding. Problem-invariant encoding representations of the categorical features, such as one-hot encoding or label encoding, can severely limit the power of prediction models. We propose deep neural networks with entity embeddings for travel mode choice prediction. We adopt the entity embedding technique to jointly learn meaningful representation of categorical variables and accurate travel mode predictions. Experiments using the London travel dataset show that deep neural networks with entity embedding technique outperform neural networks with other encoding techniques, as well as tree-based models. Besides, we found that the learned embeddings can boost the performances of tree-based models by substituting categorical features with the neural network learned features. Finally, we verify that entity embedding can learn meaningful representations of the categorical features using feature visualization at low dimensional space. INDEX TERMS Travel mode choice, entity embedding, deep neural network, machine learning.