search

thuml / SimMTM

About Code release for "SimMTM: A Simple Pre-Training Framework for Masked Time-Series Modeling" (NeurIPS 2023 Spotlight), https://arxiv.org/abs/2302.00861
107 stars 13 forks source link

experiment on cross-domain classification #1

Closed s4ndr1ver closed 10 months ago

s4ndr1ver commented 10 months ago

Hi, author! Have you noticed that in the cross-domain classification experiment, the results of the random init method are quite good? For instance, in the SleepEEG to Epilepsy classification task, the accuracy achieved by the random init method is over 95%, not like in the paper.

dongjiaxiang commented 10 months ago

Thank you for your reminder， it is so sensitive for epoch numbers and learning rates due to the unique setting (tiny training samples and large validation and test samples). We are just giving sample scripts and you can adjust the parameters appropriately and observe the effect.