23 October 2023 in IEEE Transactions on Artificial Intelligence
Multivariate long-term time-series forecasting tasks are very challenging tasks in many real-world application areas. Recently, researchers focus on designing robust and effective methods, and have made considerable progress. However, there are several issues with existing models that need to be overcome. First, the lack of a relationship structure between multivariate variables needs to be addressed. Second, most models only have a weak ability to capture local dynamic changes across the entire long-term time-series. Third, current models suffer from high computational complexity and unsatisfactory accuracy. To figure out the above issues, we propose an effective and efficient method called Multi-view Time-series Graph Structure Representation (MTGSR). MTGSR uses GCNs to construct topological relationships in the multivariate time-series from three different perspectives: time, dimension, and crossing segments. Variation trends in different dimensions are extracted through a difference operation to construct topological maps that reflect the correlations between different dimensions. To capture the dynamically changing characteristics of fluctuation correlations between adjacent local sequences, MTGSR constructs a cross graph by calculating correlation coefficients between adjacent local sequences. Extensive experiments show that MTGSR reduces errors by 17.41% over the state-of-the-art. Additionally, memory use is decreased by 66.52% and the running time is reduced by 78.09%.
23 December 2022 in IEEE Transactions on Emerging Topics in Computing
Forecasting long sequence time series plays a crucial role in many applications such as anomaly detection and financial predictions. Achieving consistently good results requires a model that can precisely capture the long-range dependencies in input sequences. And very few current models can meet the requirements. Informer has recently demonstrated state-of-the-art accuracy in LSTF. Yet several other aspects of its performance leave much room for improvement. These include: 1) complexity - Informer has a relatively high computational complexity and a high memory overhead; 2) nuance - there is limited ability to capture the subtle features in a data stream; 3) interpretability - the inference procedure of Informer is not explainable; 4) extensibility - accuracy is poor with extra-long multivariate time series. To address these issues, we propose a suite of models under the banner Sepformer. The set comprises Sepformer and two variants SWformer and Mini-SWformer. Sepformer uses separate networks to extract data stream features in parallel. SWformer and Mini-SWformer dramatically separate high-frequency and low-frequency components to process the data stream and reduce the requirement for GPU memory by adopting a discrete wavelet transform. Extensive experiments show that the Sepformer models substantially outperform state-of-the-art methods in terms of accuracy, computational complexity and usage of GPU memory use.
15 November 2022 in Future Generation Computer Systems
25 July 2022 in Neural Computing and Applications