TimeMachine: A Time Series is Worth 4 Mambas for Long-term Forecasting
Fuente:
arXiv
Saved in:
| Main Authors: | , |
|---|---|
| Format: | Preprint |
| Published: |
2024
|
| Subjects: | |
| Online Access: | |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1866916366904197120 |
|---|---|
| author | Ahamed, Md Atik Cheng, Qiang |
| author_facet | Ahamed, Md Atik Cheng, Qiang |
| contents | Long-term time-series forecasting remains challenging due to the difficulty in capturing long-term dependencies, achieving linear scalability, and maintaining computational efficiency. We introduce TimeMachine, an innovative model that leverages Mamba, a state-space model, to capture long-term dependencies in multivariate time series data while maintaining linear scalability and small memory footprints. TimeMachine exploits the unique properties of time series data to produce salient contextual cues at multi-scales and leverage an innovative integrated quadruple-Mamba architecture to unify the handling of channel-mixing and channel-independence situations, thus enabling effective selection of contents for prediction against global and local contexts at different scales. Experimentally, TimeMachine achieves superior performance in prediction accuracy, scalability, and memory efficiency, as extensively validated using benchmark datasets. Code availability: https://github.com/Atik-Ahamed/TimeMachine |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2403_09898 |
| institution | arXiv |
| publishDate | 2024 |
| record_format | arxiv |
| spellingShingle | TimeMachine: A Time Series is Worth 4 Mambas for Long-term Forecasting Ahamed, Md Atik Cheng, Qiang Machine Learning Long-term time-series forecasting remains challenging due to the difficulty in capturing long-term dependencies, achieving linear scalability, and maintaining computational efficiency. We introduce TimeMachine, an innovative model that leverages Mamba, a state-space model, to capture long-term dependencies in multivariate time series data while maintaining linear scalability and small memory footprints. TimeMachine exploits the unique properties of time series data to produce salient contextual cues at multi-scales and leverage an innovative integrated quadruple-Mamba architecture to unify the handling of channel-mixing and channel-independence situations, thus enabling effective selection of contents for prediction against global and local contexts at different scales. Experimentally, TimeMachine achieves superior performance in prediction accuracy, scalability, and memory efficiency, as extensively validated using benchmark datasets. Code availability: https://github.com/Atik-Ahamed/TimeMachine |
| title | TimeMachine: A Time Series is Worth 4 Mambas for Long-term Forecasting |
| topic | Machine Learning |
| url | https://arxiv.org/abs/2403.09898 |