Awesome
(Time Series Domain Adaptation via Sparse Associative Structure Alignment)[AAAI2021]
Requirements
-
Python 3.7
-
Pytorch 1.7
Quick Start
The pytorch version of sasa basically replicated the results of the tensorflow version of sasa on the Boiler dataset.
You can run it with the following command.
python main.py
Experiment Result
AUC scores on the Boiler data set.
| 1->2 | 1->3 | 2>1 | 2->3 | 3->1 | 3->2 | mean |
|---|---|---|---|---|---|---|
| 0.7260 | 0.9599 | 0.8575 | 0.9451 | 0.9175 | 0.6835 | 0.8483 |