As an important part of the intelligent transportation system (ITS), short-term traffic prediction has become a hot research topic in the field of traffic engineering. In recent years, with the emergence of rich traffic data and the development of deep learning technologies, neural networks have been widely used in short-term traffic forecasting. Among them, the Recurrent Neural Networks (RNN), especially the Long Short-Term Memory network (LSTM) shows the excellent ability of time-series tasks. To improve the prediction accuracy of the LSTM, some research uses the spatial-temporal matrix or Convolutional Neural Network (CNN) to extract the spatial features of the data for the LSTM network to use. In this article, we propose an attention CNN to predict traffic speed. The model uses three-dimensional data matrices constructed by traffic flow, speed, and occupancy. The spatial-temporal features extraction and the attention models are all performed by the convolution unit. Experiments on traffic data at 15-minute intervals show that the proposed algorithm has considerable advantages in predicting tasks compared to other commonly used algorithms, and the proposed algorithm has an improvement effect for cases with missing data. At the same time, by visualizing the weights generated by the attention model, we can see the influence of different spatial-temporal data on the forecasting task. C 2018 Computer-Aided Civil and Infrastructure Engineering.

The current study explored differences among cervical vestibular evoked myogenic potentials (cVEMP) that were evoked by CE-chirp and click and tone pip in healthy controls, and tried to explain the differences of cVEMP between the three of them. Thirty normal volunteers were used as subjects for CE-chirp and click and tone-pip (Blackman pip) stimuli. The latency of P1, N1, peak-to-peak P1-N1 amplitude, and cVEMP interaural difference were obtained and analyzed. The response rates of cVEMP were 93 % for click and 100 % for both Blackman pip and CE-chirp, respectively. The P1 and N1 latencies of cVEMP evoked by CE-chirp were the shortest, followed by click, with Blackman pip the longest (F = 6,686.852, P < 0.001). All indices of cVEMP evoked by the three stimuli showed no significant difference between the left and right ears or between genders. cVEMP responses were significantly different between the three stimuli. Compared with the currently used stimulus, CE-chirp can evoke cVEMP with shorter latencies and demonstrates increased speed and reliability.

The red coloration of apple (Malus × domestica Borkh.) is due to the accumulation of anthocyanins in the fruit peel. Light is essential for anthocyanin biosynthesis in apple. In this study, we performed a transcriptome sequencing (RNA-seq) analysis of apple fruit exposed to light after unbagging. The identified differentially expressed genes included MdBBX21, which is homologous to Arabidopsis BBX21, suggesting it may be involved in light-induced anthocyanin biosynthesis. Additionally, MdBBX21 was localized in the nucleus and its gene was expressed earlier than MdMYB1 in apple peel treated with light. Overexpressing MdBBX21 in Arabidopsis and apple calli under light increased anthocyanin accumulation. Dual-luciferase and yeast one-hybrid assays confirmed that MdBBX21 binds to the MdHY5, MdBBX20, and MdBBX22-1/2 promoters and induces expression. At the same time, MdHY5 can also activate the expression of MdBBX21. Furthermore, bimolecular fluorescence complementation and yeast two-hybrid assays demonstrated that MdBBX21 can interact with MdHY5. This interaction can significantly enhance MdMYB1 promoter activity. These findings clarify the molecular mechanism by which MdBBX21 positively regulates light-induced anthocyanin accumulation in apple.