TTC

Transformer-based TDE Classifier (TTC) is a time-domain classification algorithm for tidal disruption events (TDEs), based on Mgformer. Related work has been accepted by ApJ (https://doi.org/10.3847/1538-4357/ae3153).

The TTC consists of two parts: function Fitting (Curved-fitting Filter,CFF) and algorithm scoring (Transformer ML Filter).

Training_Data_Processing.ipynb: Contains the preprocessing pipeline for preparing the training dataset.

Local_Filter.ipynb: Includes the experimental results and relevant analysis associated with this work.

Please note that WFST-related data is currently not publicly available. As a result, several key directories and methods have been removed from this repository.

⚠️ This means the code may not run out of the box without additional data or reconstruction of specific components.

We appreciate your understanding. Feel free to open an issue if you have any questions or need assistance preparing your own dataset for testing.

Acknowledgements

This project is built upon Mgformer. We gratefully acknowledge their contributions.

The source code or model has been adapted from the following repository: (https://github.com/LiYunxiaoboy/mgformer)