Dongfeng Yuan

Daojun Liang, Haixia Zhang, Dongfeng Yuan et al.

As Transformer-based models have achieved impressive performance on various time series tasks, Long-Term Series Forecasting (LTSF) tasks have also received extensive attention in recent years. However, due to the inherent computational complexity and long sequences demanding of Transformer-based methods, its application on LTSF tasks still has two major issues that need to be further investigated: 1) Whether the sparse attention mechanism designed by these methods actually reduce the running time on real devices; 2) Whether these models need extra long input sequences to guarantee their performance? The answers given in this paper are negative. Therefore, to better copy with these two issues, we design a lightweight Period-Attention mechanism (Periodformer), which renovates the aggregation of long-term subseries via explicit periodicity and short-term subseries via built-in proximity. Meanwhile, a gating mechanism is embedded into Periodformer to regulate the influence of the attention module on the prediction results. Furthermore, to take full advantage of GPUs for fast hyperparameter optimization (e.g., finding the suitable input length), a Multi-GPU Asynchronous parallel algorithm based on Bayesian Optimization (MABO) is presented. MABO allocates a process to each GPU via a queue mechanism, and then creates multiple trials at a time for asynchronous parallel search, which greatly reduces the search time. Compared with the state-of-the-art methods, the prediction error of Periodformer reduced by 13% and 26% for multivariate and univariate forecasting, respectively. In addition, MABO reduces the average search time by 46% while finding better hyperparameters. As a conclusion, this paper indicates that LTSF may not need complex attention and extra long input sequences. The code has been open sourced on Github.

Daojun Liang, Haixia Zhang, Jing Wang et al.

In this paper, we find that existing online forecasting methods have the following issues: 1) They do not consider the update frequency of streaming data and directly use labels (future signals) to update the model, leading to information leakage. 2) Eliminating information leakage can exacerbate concept drift and online parameter updates can damage prediction accuracy. 3) Leaving out a validation set cuts off the model's continued learning. 4) Existing GPU devices cannot support online learning of large-scale streaming data. To address the above issues, we propose a novel online learning framework, Act-Now, to improve the online prediction on large-scale streaming data. Firstly, we introduce a Random Subgraph Sampling (RSS) algorithm designed to enable efficient model training. Then, we design a Fast Stream Buffer (FSB) and a Slow Stream Buffer (SSB) to update the model online. FSB updates the model immediately with the consistent pseudo- and partial labels to avoid information leakage. SSB updates the model in parallel using complete labels from earlier times. Further, to address concept drift, we propose a Label Decomposition model (Lade) with statistical and normalization flows. Lade forecasts both the statistical variations and the normalized future values of the data, integrating them through a combiner to produce the final predictions. Finally, we propose to perform online updates on the validation set to ensure the consistency of model learning on streaming data. Extensive experiments demonstrate that the proposed Act-Now framework performs well on large-scale streaming data, with an average 28.4% and 19.5% performance improvement, respectively. Experiments can be reproduced via https://github.com/Anoise/Act-Now.

Weitao Xu, Dongfeng Yuan, Liangfei Xue

With the continuous development of science and technology, the intelligent development of community system becomes a trend. Meanwhile, smart mobile devices and cloud computing technology are increasingly used in intelligent information systems; however, smart mobile devices such as smartphone and smart pad, also known as thin clients, limited by either their capacities (CPU, memory or battery) or their network resources, do not always meet users' satisfaction in using mobile services. Mobile cloud computing, in which resource-rich virtual machines of smart mobile device are provided to a customer as a service, can be terrific solution for expanding the limitation of real smart mobile device, but the resources utilization rate is low and the information cannot be shared easily. To address the problems above, this paper proposes an information system for intelligent community, which is composed of thin clients, wide band network and cloud computing servers. On one hand, the thin clients with the characteristics of energy efficiency, high robustness and high computing capacity can efficiently avoid the problems encountered in the PC architecture and mobile devices. On the other hand, the cloud computing servers in the proposed information system solve the problems of resource sharing barriers. Finally, the system is built in real environments to evaluate the performance. We deploy the proposed system in a community with more than 2000 residents, and it is demonstrated that the proposed system is robust and efficient.