Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

在使用单变量天频时间序列数据进行预测时,应该如何学习节假日特征? #402

Closed
Alamoooooooo opened this issue May 8, 2024 · 1 comment

Comments

@Alamoooooooo
Copy link

您好!
目前我在使用一个时间范围2020-2024年的天频单变量时间序列进行预测,预测目标是未来1个月的每天的数值此时间序列对节假日有明显规律(包括双休日/节日)。
我使用了ITransformer和TimesNet,并适当造了一些有关节假日特征和时间特征(年/月/日/周等)。但实际预测结果并没有显示出model学习到了任何有关节日的信息T .T,甚至不如单变量只输入时间序列值。
此前我使用同一套特征在LGB中进行预测,模型对于节日非常敏感,效果较好。

请问您有针对天频单变量时间序列进行预测时,对节假日规律进行调优的经验可以分享吗?或讨论区有大佬愿意分享想法,也感激不尽><

@wuhaixu2016
Copy link
Collaborator

wuhaixu2016 commented May 16, 2024

您好,这个问题我觉得可能取决于数据量,因为5年的数据只有1000多个时间点,所以对于timesnet/transformer这类模型来说,数据是不充足的。我觉得可以尝试使用我们最新的TimeMixer:https://github.com/thuml/Time-Series-Library/blob/main/models/TimeMixer.py

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
None yet
Projects
None yet
Development

No branches or pull requests

2 participants