MCformer: Multivariate Time Series Forecasting With Mixed-Channels Transformer
- 16 May 2024
- journal article
- research article
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Internet of Things Journal
- Vol. 11 (17) , 28320-28329
Abstract
The massive generation of time-series data by large-scale Internet of Things (IoT) devices necessitates the exploration of more effective models for multivariate time-series forecasting. In previous models, there was a predominant use of the channel dependence (CD) strategy (where each channel represents a univariate sequence). Current state-of-the-art (SOTA) models primarily rely on the channel independence (CI) strategy. The CI strategy treats channel multichannel series as separate single-channel series, expanding the data set to improve generalization performance and avoiding interchannel correlation that disrupts long-term features. However, the CI strategy faces the challenge of interchannel correlation forgetting. To address this issue, we propose an innovative Mixed Channels strategy, combining the data expansion advantages of the CI strategy with the ability to mitigate interchannel correlation forgetting. Based on this strategy, we introduce MCformer, a multivariate time-series forecasting model with mixed channel features. The model blends a specific number of channels, leveraging an attention mechanism to effectively capture interchannel correlation information when modeling long-term features. Experimental results demonstrate that the Mixed Channels strategy outperforms pure CI strategy in multivariate time-series forecasting tasks.Keywords
All Related Versions
Funding Information
- National Natural Science Foundation of China (62006110)
- Natural Science Foundation of Hunan Province (2024JJ7428, 2023JJ30518)
- Scientific Research Project of Hunan Provincial Department of Education (22C0229)
This publication has 20 references indexed in Scilit:
- Wavelet-Seq2Seq-LSTM with attention for time series forecasting of level of dams in hydroelectric power plantsEnergy, 2023
- Spatio-temporal hierarchical MLP network for traffic forecastingInformation Sciences, 2023
- Towards Spatio- Temporal Aware Traffic Time Series Forecasting2022
- A Multivariate-Time-Series-Prediction-Based Adaptive Data Transmission Period Control Algorithm for IoT NetworksIEEE Internet of Things Journal, 2021
- Principles and algorithms for forecasting groups of time series: Locality and globalityInternational Journal of Forecasting, 2021
- Learning Graph Structures With Transformer for Multivariate Time-Series Anomaly Detection in IoTIEEE Internet of Things Journal, 2021
- Multivariate time series dataset for space weather data analyticsScientific Data, 2020
- Reinforced Spatiotemporal Attentive Graph Neural Networks for Traffic ForecastingIEEE Internet of Things Journal, 2020
- DeepAR: Probabilistic forecasting with autoregressive recurrent networksInternational Journal of Forecasting, 2020
- Deep Belief Network for Meteorological Time Series Prediction in the Internet of ThingsIEEE Internet of Things Journal, 2018