HKUDS / FlashST

[ICML'2024] "FlashST: A Simple and Universal Prompt-Tuning Framework for Traffic Prediction"
https://arxiv.org/abs/2405.17898
Apache License 2.0
59 stars 2 forks source link

Reproduction of Results in Table4 #6

Open louisechen520 opened 1 month ago

louisechen520 commented 1 month ago

Hi, thank you for the interesting work. I have a problem, on the enhancement of STGCN, I got the result below: 2024-09-09 09:39: Average Horizon, MAE: 1.84, RMSE: 2.71, MAPE: 52.5095% The MAPE is even worse than the original STGCN, which is 50.06%, should I adjust the parameter or something to make it better.

LZH-YS1998 commented 1 month ago

Hello, thank you for your attention. For more accurate results, please set the OutputLayer parameter to True in the STGCN configuration within FlashST.py. Thank you.