Zhiqiang Zhang
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Zhiqiang Zhang has a total of 18 co-authors. Here are the top co-authors by number of shared publications:
- Dandan Zhang
Key Laboratory of Computer Network and Information Integration, Ministry of Education, School of Computer Science and Engineering, Southeast University, Nanjing, China
2 shared publications - Yun Wang
School of Computer Science and Engineering, Southeast University, Nanjing 211189, China
2 shared publications - Binbin Hu
Ant Group, Hangzhou, China
1 shared publication Dandan Zhang
1 shared publication- Hengbin Cui
Alibaba Group, Hangzhou, China
1 shared publication - Hongbing Wang
Guizhou Normal University
1 shared publication
Latest Publications
- 9 October 2025
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Knowledge and Data Engineering
- Vol. 37 (12) , 6750-6764
A Data-Driven Scale-Adaptive Time-Frequency Convolutional Network for Long Sequence Time-Series Forecasting
Models based on Transformer variants have consistently demonstrated leading performance in long sequence time series forecasting. However, in some complex application scenarios, Transformers tend to capture low-frequency information in the data while overlooking high-frequency information, which often contains rich non-stationary features. This unbalanced feature extraction approach limits the model’s ability to effectively handle real-world time series data. To address this issue, we explicitly represent both low-frequency and high-frequency information and propose a model called STCNet, a data-driven scale-adaptive convolutional network that aims to extract diverse features and patterns from the data by learning features across different frequency bands in a balanced manner. Specifically, we propose an entropy-based adaptive wavelet basis selection algorithm, which can adaptively select appropriate wavelet bases based on the data distribution to achieve effective multi-frequency decomposition of complex time series. In addition, we designed a hierarchical scale-adaptive factor that allows for dynamic adjustment of feature weights according to different time scales through refined layered weight adjustment, significantly enhancing the model’s capability in handling non-stationary time series features. To further optimize the output features of the model, we introduce a test-time training mechanism, combined with a fast weight update strategy and a weight-sharing strategy to reduce the number of model parameters, effectively mitigating the risk of overfitting. Experimental results on nine datasets demonstrate that STCNet outperforms the current state-of-the-art models in both effectiveness and efficiency.
- 9 October 2025
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Knowledge and Data Engineering
- Vol. 37 (12) , 6750-6764
A Data-Driven Scale-Adaptive Time-Frequency Convolutional Network for Long Sequence Time-Series Forecasting
Models based on Transformer variants have consistently demonstrated leading performance in long sequence time series forecasting. However, in some complex application scenarios, Transformers tend to capture low-frequency information in the data while overlooking high-frequency information, which often contains rich non-stationary features. This unbalanced feature extraction approach limits the model’s ability to effectively handle real-world time series data. To address this issue, we explicitly represent both low-frequency and high-frequency information and propose a model called STCNet, a data-driven scale-adaptive convolutional network that aims to extract diverse features and patterns from the data by learning features across different frequency bands in a balanced manner. Specifically, we propose an entropy-based adaptive wavelet basis selection algorithm, which can adaptively select appropriate wavelet bases based on the data distribution to achieve effective multi-frequency decomposition of complex time series. In addition, we designed a hierarchical scale-adaptive factor that allows for dynamic adjustment of feature weights according to different time scales through refined layered weight adjustment, significantly enhancing the model’s capability in handling non-stationary time series features. To further optimize the output features of the model, we introduce a test-time training mechanism, combined with a fast weight update strategy and a weight-sharing strategy to reduce the number of model parameters, effectively mitigating the risk of overfitting. Experimental results on nine datasets demonstrate that STCNet outperforms the current state-of-the-art models in both effectiveness and efficiency.
- 10 September 2024
- Published by Elsevier in Pattern Recognition
- Vol. 158
Dynamic convolutional time series forecasting based on adaptive temporal bilateral filtering
- 10 September 2024
- Published by Elsevier in Pattern Recognition
- Vol. 158
Dynamic convolutional time series forecasting based on adaptive temporal bilateral filtering
- 17 April 2024
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Knowledge and Data Engineering
- Vol. 36 (11) , 6181-6194
Fast Long Sequence Time-Series Forecasting for Edge Service Running State Based on Data Drift and Non-Stationarity
The operational state of edge services in remote areas and complex geographical work environments is influenced by scarce hardware resources, unstable communication state and external dynamic environment. This leads to the continuous evolution of reliability data streams of edge services in the form of data drift, producing non-stationary data streams with substantial amounts of noise data, resulting in the difficulty of service operational status prediction and low efficiency. Previous studies mainly utilize stationarity-based methods to attenuate the non-stationarity of original sequence, aiming to enhance predictability using deep learning models. However, stationary series deprived of their inherent non-stationarity struggle to accurately forecast practical emergencies. Furthermore, the size and computational complexity of deep learning models are unsuitable for deployment at the edge. To accurately predict the reliability of edge services in dynamic environments, we propose a lightweight and fast long sequence time-series forecasting method based on data drift and non-stationarity, named FSNet. FSNet introduces a non-stationary information sampling factor to extract external factors influencing data flows and incorporates Moore-Penrose inverse matrix to swiftly update the model weights during runtime. Combining unmanned aerial vehicles as mobile edge servers with edge computing offloading achieves collaborative computation of the model, thereby alleviating the scarcity of resources at the edge, enhancing computational efficiency, and enabling fast and reliable prediction of service operations. Extensive experimental results validate the effectiveness and efficiency of the FSNet.
- 17 April 2024
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Knowledge and Data Engineering
- Vol. 36 (11) , 6181-6194
Fast Long Sequence Time-Series Forecasting for Edge Service Running State Based on Data Drift and Non-Stationarity
The operational state of edge services in remote areas and complex geographical work environments is influenced by scarce hardware resources, unstable communication state and external dynamic environment. This leads to the continuous evolution of reliability data streams of edge services in the form of data drift, producing non-stationary data streams with substantial amounts of noise data, resulting in the difficulty of service operational status prediction and low efficiency. Previous studies mainly utilize stationarity-based methods to attenuate the non-stationarity of original sequence, aiming to enhance predictability using deep learning models. However, stationary series deprived of their inherent non-stationarity struggle to accurately forecast practical emergencies. Furthermore, the size and computational complexity of deep learning models are unsuitable for deployment at the edge. To accurately predict the reliability of edge services in dynamic environments, we propose a lightweight and fast long sequence time-series forecasting method based on data drift and non-stationarity, named FSNet. FSNet introduces a non-stationary information sampling factor to extract external factors influencing data flows and incorporates Moore-Penrose inverse matrix to swiftly update the model weights during runtime. Combining unmanned aerial vehicles as mobile edge servers with edge computing offloading achieves collaborative computation of the model, thereby alleviating the scarcity of resources at the edge, enhancing computational efficiency, and enabling fast and reliable prediction of service operations. Extensive experimental results validate the effectiveness and efficiency of the FSNet.
- 17 August 2023
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Neural Networks and Learning Systems
- Vol. 35 (11) , 16368-16382
CTFNet: Long-Sequence Time-Series Forecasting Based on Convolution and Time–Frequency Analysis
Although current time-series forecasting methods have significantly improved the state-of-the-art (SOTA) results for long-sequence time-series forecasting (LSTF), they still have difficulty in capturing and extracting the features and dependencies of long-term sequences and suffer from information utilization bottlenecks and high-computational complexity. To address these issues, a lightweight single-hidden layer feedforward neural network (SLFN) combining convolution mapping and time–frequency decomposition called CTFNet is proposed with three distinctive characteristics. First, time-domain (TD) feature mining—in this article, a method for extracting the long-term correlation of horizontal TD features based on matrix factorization is proposed, which can effectively capture the interdependence among different sample points of a long time series. Second, multitask frequency-domain (FD) feature mining—this can effectively extract different frequency feature information of time-series data from the FD and minimize the loss of data features. Integrating multiscale dilated convolutions, simultaneously focusing on both global and local context feature dependencies at the sequence level, and mining the long-term dependencies of the multiscale frequency information and the spatial dependencies among the different scale frequency information, break the bottleneck of data utilization, and ensure the integrity of feature extraction. Third, highly efficient—the CTFNet model has a short training time and fast inference speed. Our empirical studies with nine benchmark datasets show that compared with state-of-the-art methods, CTFNet can reduce prediction error by 64.7% and 53.7% for multivariate and univariate time series, respectively.
- 17 August 2023
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Neural Networks and Learning Systems
- Vol. 35 (11) , 16368-16382
CTFNet: Long-Sequence Time-Series Forecasting Based on Convolution and Time–Frequency Analysis
Although current time-series forecasting methods have significantly improved the state-of-the-art (SOTA) results for long-sequence time-series forecasting (LSTF), they still have difficulty in capturing and extracting the features and dependencies of long-term sequences and suffer from information utilization bottlenecks and high-computational complexity. To address these issues, a lightweight single-hidden layer feedforward neural network (SLFN) combining convolution mapping and time–frequency decomposition called CTFNet is proposed with three distinctive characteristics. First, time-domain (TD) feature mining—in this article, a method for extracting the long-term correlation of horizontal TD features based on matrix factorization is proposed, which can effectively capture the interdependence among different sample points of a long time series. Second, multitask frequency-domain (FD) feature mining—this can effectively extract different frequency feature information of time-series data from the FD and minimize the loss of data features. Integrating multiscale dilated convolutions, simultaneously focusing on both global and local context feature dependencies at the sequence level, and mining the long-term dependencies of the multiscale frequency information and the spatial dependencies among the different scale frequency information, break the bottleneck of data utilization, and ensure the integrity of feature extraction. Third, highly efficient—the CTFNet model has a short training time and fast inference speed. Our empirical studies with nine benchmark datasets show that compared with state-of-the-art methods, CTFNet can reduce prediction error by 64.7% and 53.7% for multivariate and univariate time series, respectively.
- 1 April 2021
- Published by Institute of Electrical and Electronics Engineers (IEEE)
- Vol. 30, 2559-2570
IntelliTag: An Intelligent Cloud Customer Service System Based on Tag Recommendation
To reduce the customer service pressure of small and medium-sized enterprises, we propose an intelligent cloud customer service system, called IntelliTag. Unlike traditional customer service, a cloud service based system has difficulty in collecting user personal information. Therefore, we add a tag recommendation function to quickly capture the user’s question intent by clicking on the tags. Specifically, IntelliTag is elaborately designed with the consideration of the following three aspects. First, how to mine high-quality tags is a challenging problem. Second, in the tag recommendation tasks, we have multifarious data types and relations that are used to build a sequential recommendation model. Finally, system implementation and deployment also need to be carefully designed to satisfy online service requirements. In this paper, we show the details of data construction, model designs, system implementation and deployment, and the empirical results compared with several state-of-the-art methods. Nowadays, our IntelliTag has already supported hundreds of thousands of enterprises and millions of users in our industrial production environment.
- 1 April 2021
- Published by Institute of Electrical and Electronics Engineers (IEEE)
- Vol. 30, 2559-2570
IntelliTag: An Intelligent Cloud Customer Service System Based on Tag Recommendation
To reduce the customer service pressure of small and medium-sized enterprises, we propose an intelligent cloud customer service system, called IntelliTag. Unlike traditional customer service, a cloud service based system has difficulty in collecting user personal information. Therefore, we add a tag recommendation function to quickly capture the user’s question intent by clicking on the tags. Specifically, IntelliTag is elaborately designed with the consideration of the following three aspects. First, how to mine high-quality tags is a challenging problem. Second, in the tag recommendation tasks, we have multifarious data types and relations that are used to build a sequential recommendation model. Finally, system implementation and deployment also need to be carefully designed to satisfy online service requirements. In this paper, we show the details of data construction, model designs, system implementation and deployment, and the empirical results compared with several state-of-the-art methods. Nowadays, our IntelliTag has already supported hundreds of thousands of enterprises and millions of users in our industrial production environment.