二维表示的 Time Analysis 框架 — TimesNet

傅科摆 _ py

已于 2025-03-21 19:46:46 修改

阅读量1.3k

点赞数 26

分类专栏：文章解读文章标签：人工智能算法

于 2024-11-03 11:57:08 首次发布

本文链接：https://blog.youkuaiyun.com/m0_60890610/article/details/143462880

版权

出处：ICLR 2023

代码：https://github.com/thuml/TimesNet

贡献

1. Motivated by multi-periodicity and complex interactions within and between periods, we find out a modular way for temporal variation modeling. By transforming the 1D time series into 2D space, we can present the intraperiod- and interperiod-variations simultaneously.
2. We propose the TimesNet with TimesBlock to discover multiple periods and capture temporal 2D-variations from transformed 2D tensors by a parameter-efficient inception block.
3. As a task-general foundation model, TimesNet achieves the consistent state-of-the-art in five mainstream time series analysis tasks. Detailed and insightful visualizations are included.