Search

In natural language processing (NLP), transformer based models have surpassed recurrent neural networks (RNN) as state of the art, being introduced specifically to address the limitations of RNNs originating from its sequential nature. As a similar sequence modeling problem, transformer methods can be readily adapted for deep learning time series prediction. This paper proposes a sparse transformer based approach for electricity load prediction. The layers of a transformer addresses the shortcomings of RNNs and CNNs by applying the attention mechanism on the entire time series, allowing any data point in the input to influence any location in the output of the layer. This allows transformers to incorporate information from the entire sequence in a single layer. Attention computations can also be parallelized. Thus, transformers can achieve faster speeds, or trade this speed for more layers and increased complexity. In experiments on public datasets, the sparse transformer attained comparable accuracy to an RNN-based SOTA method (Liu et al., 2022) while being up to 5× faster during inference. Moreover, the proposed model is general enough to forecast the load from individual households to city levels as shown in the extensive experiments conducted. [ABSTRACT FROM AUTHOR]

In recent years, with the development of computer science, deep learning is held as competent enough to solve the problem of inference and learning in high dimensional space. Therefore, it has received unprecedented attention from both the academia and the business community. Compared with CPU/GPU, FPGA has attracted much attention for its high-energy efficiency, short development cycle and reconfigurability in the aspect of deep learning algorithm. However, because of the limited research on OpenCL optimization on FPGA of deep learning algorithms, OpenCL tools and models applied to CPU/GPU cannot be directly used on FPGA. This makes it difficult for software programmers to use FPGA when implementing deep learning algorithms for a rewarding performance. To solve this problem, this paper proposed an OpenCL computational model based on FPGA template architecture to optimize the time-consuming convolution layer in deep learning. The comparison between the program applying the computational model and the corresponding optimization program provided by Xilinx indicates that the former is 8-40 times higher than the latter in terms of performance. [ABSTRACT FROM AUTHOR]

Embedding network data into a low-dimensional vector space has shown promising performance for many real-world applications, such as node classification and entity retrieval. However, most existing methods focused only on leveraging network structure. For social networks, besides the network structure, there also exists rich information about social actors, such as user profiles of friendship networks and textual content of citation networks. These rich attribute information of social actors reveal the homophily effect, exerting huge impacts on the formation of social networks. In this paper, we explore the rich evidence source of attributes in social networks to improve network embedding. We propose a generic Attributed Social Network Embedding framework (ASNE), which learns representations for social actors (i.e., nodes) by preserving both thestructural proximityandattribute proximity. While thestructural proximitycaptures the global network structure, theattribute proximityaccounts for the homophily effect. To justify our proposal, we conduct extensive experiments on four real-world social networks. Compared to the state-of-the-art network embedding approaches,ASNEcan learn more informative representations, achieving substantial gains on the tasks of link prediction and node classification. Specifically,ASNEsignificantly outperformsnode2vecwith an 8.2 percent relative improvement on the link prediction task, and a 12.7 percent gain on the node classification task. [ABSTRACT FROM AUTHOR]